<!DOCTYPE html><html lang="en" xmlns="http://www.w3.org/1999/xhtml" xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" style="font-size:16px;"><head></head><head><meta charset="utf-8"/><!--[if !mso]><!--><meta http-equiv="X-UA-Compatible" content="IE=edge"/><!--<![endif]--><meta name="viewport" content="width=device-width,initial-scale=1"/><meta name="x-apple-disable-message-reformatting"/><meta name="format-detection" content="telephone=no,address=no,email=no,date=no,url=no"/><meta name="color-scheme" content="light"/><meta name="supported-color-schemes" content="light"/><title>Hogwild! Inference: Parallel LLM Generation via Concurrent Attention </title><!--[if mso]><xml><o:OfficeDocumentSettings><o:AllowPNG/><o:PixelsPerInch>96</o:PixelsPerInch></o:OfficeDocumentSettings></xml><![endif]--><style>
:root { color-scheme: light; supported-color-schemes: light; }
body { margin: 0; padding: 0; min-width: 100%!important; -ms-text-size-adjust: 100% !important; -webkit-transform: scale(1) !important; -webkit-text-size-adjust: 100% !important; -webkit-font-smoothing: antialiased !important; }
.body { word-wrap: normal; word-spacing:normal; }
table.mso { width: 100%; border-collapse: collapse; padding: 0; table-layout: fixed; }
img { border: 0; outline: none; }
table { mso-table-lspace: 0px; mso-table-rspace: 0px; }
td, a, span { mso-line-height-rule: exactly; }
#root [x-apple-data-detectors=true],
a[x-apple-data-detectors=true],
#MessageViewBody a { color: inherit !important; text-decoration: inherit !important; font-size: inherit !important; font-family: inherit !important; font-weight: inherit !important; line-height: inherit !important; }
span.MsoHyperlink { color: inherit !important; mso-style-priority: 99 !important; }
span.MsoHyperlinkFollowed { color: inherit !important; mso-style-priority: 99 !important; }
.a { background-color:#dedede; }
.b { background-color:#2a2a2a; }
.c { background-color:#ffffff; }
.d { background-color:#fff0c8; }
.d2 { background-color:#FFFFFF; }
.d3 { background-color:#FFFFFF; }
h1 a { text-decoration:none;color:#2C81E5;font-style:italic; }
h2 a { text-decoration:none;color:#2C81E5;font-style:italic; }
h3 a { text-decoration:none;color:#2C81E5;font-style:italic; }
h4 a { text-decoration:none;color:#2C81E5;font-style:italic; }
h5 a { text-decoration:none;color:#2C81E5;font-style:italic; }
h6 a { text-decoration:none;color:#2C81E5;font-style:italic; }
h1, h1 a, h2, h2 a, h3, h3 a, h4, h4 a, h5, h5 a, h6, h6 a, ul, li, ol, p, p a { margin: 0;padding: 0; }
h1 { font-family:'Trebuchet MS','Lucida Grande',Tahoma,sans-serif;font-weight:700;font-size:28px;color:#2A2A2A;line-height:42px;padding-bottom:4px;padding-top:16px;mso-margin-top-alt:16px;mso-margin-bottom-alt:4px }
h2 { font-family:'Trebuchet MS','Lucida Grande',Tahoma,sans-serif;font-weight:700;font-size:24px;color:#2A2A2A;line-height:36px;padding-bottom:4px;padding-top:16px;mso-margin-top-alt:16px;mso-margin-bottom-alt:4px }
h3 { font-family:'Trebuchet MS','Lucida Grande',Tahoma,sans-serif;font-weight:400;font-size:20px;color:#2A2A2A;line-height:30px;padding-bottom:4px;padding-top:16px;mso-margin-top-alt:16px;mso-margin-bottom-alt:4px }
h4 { font-family:'Trebuchet MS','Lucida Grande',Tahoma,sans-serif;font-weight:400;font-size:18px;color:#2A2A2A;line-height:27px;padding-bottom:4px;padding-top:16px;mso-margin-top-alt:16px;mso-margin-bottom-alt:4px }
h5 { font-family:'Trebuchet MS','Lucida Grande',Tahoma,sans-serif;font-weight:400;font-size:16px;color:#2A2A2A;line-height:24px;padding-bottom:4px;padding-top:16px;mso-margin-top-alt:16px;mso-margin-bottom-alt:4px }
h6 { font-family:'Trebuchet MS','Lucida Grande',Tahoma,sans-serif;font-weight:400;font-size:14px;color:#2A2A2A;line-height:21px;padding-bottom:4px;padding-top:16px;mso-margin-top-alt:16px;mso-margin-bottom-alt:4px }
p { font-family:'Georgia','Times New Roman',serif;font-weight:400;color:#2D2D2D;font-size:16px;line-height:24px;padding-bottom:8px;padding-top:8px;mso-margin-top-alt:8px;mso-margin-bottom-alt:8px; }
p a, .e a, ul a, li a, .h a, .h2 a, .h3 a { word-break:break-word;color:#2C81E5 !important;text-decoration:none;font-style:italic; }
p a span, .e a span, ul a span, li a span { color: inherit }
p .bold { font-weight:bold;color:#2D2D2D; }
p span[style*="font-size"] { line-height: 1.6; }
.f p { font-size:12px;line-height:15px;color:#2D2D2D;padding:0; }
.f p a { color:#2D2D2D !important; }
.g p { font-family:'Helvetica',Arial,sans-serif;font-size:14px;line-height:20px;font-weight:normal;margin:0; }
.g p a { text-decoration: underline; }
.i p { font-family:'Helvetica',Arial,sans-serif;line-height:23px;font-size:15px;color:#2D2D2D; }
.i p a { color:#2D2D2D !important; }
.i2 p { font-family:'Helvetica',Arial,sans-serif;line-height:23px;font-size:15px;color:#2D2D2D; }
.i2 p a { color:#2D2D2D !important; }
.i3 p { font-family:'Helvetica',Arial,sans-serif;line-height:43px;font-size:24px;color:#2D2D2D; }
.i3 p a { color:#2D2D2D !important; }
.h p a { color:#595959 !important; }
.h2 p a { color:#595959 !important; }
.h3 p a { color:#595959 !important; }
.f p a, .i p a, .i2 p a, .i3 p a, .h p a, .h2 p a, .h3 p a { text-decoration:underline; }
.j { border-top:3px solid #ffeb2d; }
.k p { padding-left:15px;padding-bottom:0px;padding-top:6px;mso-margin-top-alt:6px;mso-margin-bottom-alt:0px;mso-margin-left-alt:15px; }
.o { background-color:#FFFFFF;border:1px solid #F1F1F1;border-radius:5px; }
.o p { font-family:'Helvetica',Arial,sans-serif;padding:0px;margin:0px; }
.l p,
.l p a { font-size:14px;line-height:20px;font-weight: bold;color:#2D2D2D;padding-bottom:6px;mso-margin-bottom-alt:6px;text-decoration:none; }
.m p,
.m p a { font-size:13px;line-height:18px;font-weight:400;color:#2D2D2D;padding-bottom:6px;mso-margin-bottom-alt:6px;text-decoration:none; }
.n p,
.n p a { font-size:12px;line-height:17px;font-weight:400;color:#2D2D2D;padding-bottom:6px;mso-margin-bottom-alt:6px;text-decoration:none; }
.p { background-color:#FFFFFF;max-width:520px;border:1px solid #E1E8ED;border:1px solid rgba(80, 80, 80, 0.3);border-radius:5px; }
.q { font-size:16px;font-family:Helvetica,Roboto,Calibri,sans-serif !important;border:1px solid #e1e8ed;border:1px solid rgba(80, 80, 80, 0.3);border-radius:10px;background-color:#FFFFFF; }
.q p { font-size:16px;font-family:system-ui,Helvetica,Roboto,Calibri,sans-serif !important;color:#222222;padding:4px 0; }
.r { border:1px solid #E1E8ED !important;border-radius:5px; }
.s p { font-size: 14px; line-height: 17px; font-weight: 400; color: #697882; text-decoration: none; }
.t p { font-family:'Helvetica',Arial,sans-serif;font-size:12px;line-height:18px;font-weight:400;color:#000000;font-style:italic;padding:0; }
.v { border-radius:10px;border:solid 0px #DFD150;background-color:#2C81E5;font-family:'Open Sans','Segoe UI','Apple SD Gothic Neo','Lucida Grande','Lucida Sans Unicode',sans-serif;color:#FFFFFF; }
.v a { text-decoration:none;display:block;color:#FFFFFF; }
.w p { font-size:12px;line-height:15px;font-weight:400;color:#FFFFFF; }
.w p a { text-decoration: underline !important;color:#FFFFFF !important; }
ul { font-family:'Helvetica',Arial,sans-serif;margin:0px 0px 0px 25px !important;padding:0px !important;color:#2D2D2D;line-height:24px;list-style:disc;font-size:16px; }
ul > li { font-family:'Helvetica',Arial,sans-serif;margin:10px 0px 0px 0px !important;padding: 0px 0px 0px 0px !important; color: #2D2D2D; list-style:disc; }
ol { font-family:'Helvetica',Arial,sans-serif;margin: 0px 0px 0px 25px !important;padding:0px !important;color:#2D2D2D;line-height:24px;list-style:decimal;font-size:16px; }
ol > li { font-family:'Helvetica',Arial,sans-serif;margin:10px 0px 0px 0px !important;padding: 0px 0px 0px 0px !important; color: #2D2D2D; list-style:decimal; }
.e h3,
.e p,
.e span { padding-bottom:0px;padding-top:0px;mso-margin-top-alt:0px;mso-margin-bottom-alt:0px; }
.e span,
.e li { font-family:'Helvetica',Arial,sans-serif;font-size:16px;color:#2D2D2D;line-height:24px; }
.rec { font-family: ui-sans-serif, system-ui, -apple-system, BlinkMacSystemFont, "Segoe UI", Roboto, "Helvetica Neue", Arial, "Noto Sans", sans-serif, "Apple Color Emoji", "Segoe UI Emoji", "Segoe UI Symbol", "Noto Color Emoji" !important; }
.rec__button:hover { background-color: #f9fafb !important; }
.copyright a {color: inherit !important; text-decoration: none !important; font-size: inherit !important; font-family: inherit !important; font-weight: inherit !important; line-height: inherit !important;}
.txt_social p { padding: 0; word-break: break-all; }
.table, .table-c, .table-h { border: 1px solid #C0C0C0; }
.table-c { padding:5px; background-color:#FFFFFF; }
.table-c p { color: #2D2D2D; font-family:'Helvetica',Arial,sans-serif !important;overflow-wrap: break-word; }
.table-h { padding:5px; background-color:#F1F1F1; }
.table-h p { color: #2A2A2A; font-family:'Trebuchet MS','Lucida Grande',Tahoma,sans-serif !important;overflow-wrap: break-word; }
@media only screen and (max-width:667px) {
.aa { width: 100% !important; }
.bb img { width: 100% !important; height: auto !important; max-width: none !important; }
.cc { padding: 0px 8px !important; }
.ee { padding-top:10px !important;padding-bottom:10px !important; }
.ff ul, .ff ol { margin: 0px 0px 0px 10px !important;padding: 0px !important; }
.ff li { margin:10px 0px 0px 10px !important; }
.r {height:140px !important;}
.s p { font-size:13px !important;line-height:15px !important; }
.mob-hide {display:none !important;}
.mob-stack {display:block !important;width:100% !important;}
.mob-w-full {width:100% !important;}
.mob-block {display:block !important;}
.embed-img {padding:0px 0px 12px 0px !important;}
.socialShare {padding-top:15px !important;}
.rec { padding-left:15px!important;padding-right:15px!important; }
.bodyWrapper { padding:7px 4px 7px 4px !important; }
.social-mobile {float:left !important;margin-top:10px !important;}
}
@media screen and (max-width: 480px) {
u + .a .gg { width: 100% !important; width: 100vw !important; }
.tok-heart { padding-top:75% !important; }
.tok-play { padding-top: 250px !important; }
}
@media screen and (max-width: 320px) {
.tok-heart { padding-top:65% !important; }
}
.u { border: 1px solid #CACACA !important; border-radius: 2px !important; background-color: #ffffff !important; padding: 0px 13px 0px 13px !important; font-family:ui-sans-serif,system-ui,-apple-system,BlinkMacSystemFont,"Segoe UI",Roboto,"Helvetica Neue",Arial,"Noto Sans",sans-serif !important;font-size: 12px !important; color: #767676 !important; }
.u a { text-decoration: none; display: block !important; color: #767676 !important; margin: 0px !important; }
.u span, .u img { color: #767676 !important;margin:0px !important; max-height:32px !important;background-color:#ffffff !important; }
</style><!--[if mso]><style type="text/css">
sup { font-size: 100% !important;vertical-align: .5em !important;mso-text-raise: -1.5% !important;line-height: 0 !important; }
ul { margin-left:0px !important; margin-right:10px !important; margin-top:20px !important; margin-bottom:20px !important; }
ul li { margin-left: 0px !important; mso-special-format: decimal; }
ol { margin-left:0px !important; margin-right:10px !important; margin-top:20px !important; margin-bottom:20px !important; }
ol li { margin-left: 0px !important; mso-special-format: decimal; }
li.listItem { margin-left:15px !important; margin-top:0px !important; }
.paddingDesktop { padding: 10px 0 !important; }
.edm_outlooklist { margin-left: -20px !important; }
.embedImage { display:none !important; }
</style><![endif]--><style>
@font-face {
font-family: 'Open Sans';
font-style: normal;
font-weight: 700;
font-display: swap;
src: url('https://fonts.gstatic.com/s/opensans/v40/memSYaGs126MiZpBA-UvWbX2vVnXBbObj2OVZyOOSr4dVJWUgsg-1x4gaVIUwaEQbjA.woff2') format('woff2');
}
@font-face {
font-family: 'Open Sans';
font-style: italic;
font-weight: 700;
font-display: swap;
src: url('https://fonts.googleapis.com/css2?family=Open+Sans:ital,wght@1,700&display=swap') format('woff2');
}
</style></head><body class="a" style="margin:0px auto;padding:0px;word-wrap:normal;word-spacing:normal;background-color:#dedede;"><div role="article" aria-roledescription="email" aria-label="email_name" lang="en" style="font-size:1rem"><div style="display:none;max-height:0px;overflow:hidden;"> Plus more about One-Minute Video Generation with Test-Time Training and Gaussian Mixture Flow Matching Models  ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ </div><table role="none" width="100%" border="0" cellspacing="0" align="center" cellpadding="0" class="gg"><tr><td align="center" valign="top"><table role="none" width="670" border="0" cellspacing="0" cellpadding="0" class="aa" style="width:670px;table-layout:fixed;"><tr><td class="bodyWrapper" align="center" valign="top" style="padding:7px 7px 7px 7px;"><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0" align="center"><tr><td align="center" valign="top" style="border-width:0px 0px 0px 0px;border-style: solid; border-color: #2a2a2a;border-radius:10px 10px 0px 0px;background-color:#ffffff;" class="c"><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0" align="center"><tr id="header"><td style="padding:28px 28px 0 28px;"><div style="padding-top:0px;padding-right:0px;padding-bottom:20px;padding-left:0px;"><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0" align="center"><tr><td class="f" align="right" valign="top"><p> April 15, 2025 | <a href="https://elink4f7.mail.bycloud.ai/ss/c/u001.c6q0w4g5sodbtO4I1B_pxSdB5RCIH6yy1Fm1CYma3Eyn2tQIa2bgAd3bk0wfzsyaSLpc6hZAnumYalGk9m-lST4p8pLQIdoFo9U_pR9UowqXjAY-dtx1TjHwRZMjZ7HwNQU_RZi0iIF80K3Nr6SiiihY8vaDK4ru3rhvvK9bBYvW782_dqjmu5Y5QQA4ohQ8b6CWeUKQvD7BbQkMaOqqOEffBur3sPm_bHsXmFeo843JLRqD4XJjdlnFXd8ZEhHuXXbT2gMfl2TJhQO_v1FeiWd1HvR3L0v2UT3DvQ6QmAmR-gQW252bqaS4tfyHIg-fKoPqmAi3f8EPuYjS8f-NxCfCRddxXixhCLzohLJmAuyciLTsbsh1lPG1EEtlOrWG3miIzSotISs6dDbksUdiVxpHV2abUj5SBNXkr1sLJFxfZSz-eUNqAG-zM7N_FUzk4rnuSSDTlp-ZT5X17DB8Gai5qKBQJJkABqg6-VJMYIr8gm2VL3PAodGfNAdqzGlqWcKVY9Jh7rQbP0a_hkE-sCoWxwGSxhk1783yMVMCHFJIF_tY_6e-rwR4xYBf2APRmL0ghKuRPGcz1BrfWfc-mWIht1oGa6gkFG_X5YGoei5VH08ISa5_EvPixvoaLgCKMbRlN1OgJLXxPix24756adQXiT8xRmXqhneRtK46p9LzfLqy6dKQtxu4TlKGCze-Lc-Zg58d17smJ5kexxqyDw/4fo/GDpba6SsTIiMseR72g6Sfw/h0/h001.DZEhLYpBKhja_ajP-ME1IV51WV_HdbEaN0LcHrAmWWc"><span class="translation_missing" title="translation missing: en.templates.posts.email.v3.header.read_online">Read Online</span></a></p></td></tr><tr><td class="dd" align="center" valign="top" style="padding:15px 0;"><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0" align="center"><tr><td align="center" valign="top"><h1 style="text-align:left;font-family:'Open Sans','Segoe UI','Apple SD Gothic Neo','Lucida Grande','Lucida Sans Unicode',sans-serif;font-weight:Bold;font-size:32px;color:#2A2A2A;padding:2px 0;line-height:38px;"> Hogwild! Inference: Parallel LLM Generation via Concurrent Attention </h1><p style="text-align:left;font-family:'Helvetica',Arial,sans-serif;font-weight:normal;font-size:20px;color:#3E3E3E;padding:5px 0;line-height:24px;"> Plus more about One-Minute Video Generation with Test-Time Training and Gaussian Mixture Flow Matching Models </p></td></tr></table></td></tr><tr><td style="height:0px;width:0px;"><div style="height:1px;" data-open-tracking="true"> <img src="https://elink4f7.mail.bycloud.ai/ss/o/u001.3wmUuY8gEWd4_869a_eXcg/4fo/GDpba6SsTIiMseR72g6Sfw/ho.gif" alt="" width="1" height="1" border="0" style="height:1px !important;width:1px !important;border-width:0 !important;margin-top:0 !important;margin-bottom:0 !important;margin-right:0 !important;margin-left:0 !important;padding-top:0 !important;padding-bottom:0 !important;padding-right:0 !important;padding-left:0 !important;"/> </div></td></tr></table></div></td></tr><tr id="content-blocks"><td class="email-card-body" align="center" valign="top" style="padding-bottom:28px;"><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0" align="center"><tr><td id="nov-18-th-nov-24-th-33-latest-ai-re" class="dd" align="left" valign="top" style="color:#2A2A2A;font-weight:normal;padding:0px 28px;text-align:left;"><h6 style="color:#2A2A2A;font-weight:normal;"><i>Apr 7th ~ Apr 13th</i><br><i>#51 Latest AI Research Explained Simply</i></h6></td></tr><tr><td><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0" style=""><tr><td bgcolor="#222222" style="background-color:#222222;padding:0.0px 0.0px 0.0px 0.0px;"><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0"><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"></p></td></tr></table></td></tr></table></td></tr><tr><td id="industry-news-in-1-line" class="dd" align="left" valign="top" style="color:#2A2A2A;font-weight:Bold;padding:0px 28px;text-align:left;"><h2 style="color:#2A2A2A;font-weight:Bold;">🗞️ Industry News in 1 Line</h2></td></tr><tr><td style="padding-bottom:12px;padding-left:50px;padding-right:40px;padding-top:12px;" class="ee"><div style="margin-left:0px;" class="edm_outlooklist"><ol start="1" style="list-style-type:decimal;margin:0px 0px;padding:0px 0px 0px 0px;"><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"><span style="background-color:#e0e0e0;"><span style="color:rgb(255, 58, 58);font-size:0.6rem;">♥ 3.9k</span></span> OpenAI has announced GPT-4.1, a next-gen non-reasoning multimodal model. GPT-4.1 is the first ever OpenAI model series to offer up to <b>1 million tokens</b>. These models, are <i>only</i> accessible via the OpenAI API, while GPT-4.5 Preview will be deprecated. </p><table role="none" border="0" cellspacing="0" cellpadding="0" style="margin:0 auto 0 auto;"><tr><td align="center" valign="top" style="width:450px;"><a href="https://elink4f7.mail.bycloud.ai/ss/c/u001.1r-Oo1v2APbJapoJieKk3bRUQFOFWXnb20kvcLCdB2VIJDskJA0rPyYEWthPs8EoIG2dPtI8pohtERVIio1cNo65MH7fd26iYxQXtijmnT3ni5lGRfRjvJUwxJv2IwtTkXjbW2jiFlpkOGNwUb51n6nfLgvfTHRG38TagizBLc6-97ia21qY3gg5UPKSr_TW/4fo/GDpba6SsTIiMseR72g6Sfw/h1/h001.eATjYCdEc9BbHjVmgJ_XLonwX0vswLCWno2v1qL_h9c" target="_blank" rel="noopener noreferrer nofollow" style="text-decoration:none;"><img src="https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/b56d6cde-fe16-4bd4-8aaa-80221107ed7c/GohELHLWoAA1_zh.jpg?t=1744740360" alt="" height="auto" width="450" style="display:block;width:100%;" border="0"/></a></td></tr><tr><td align="center" valign="top" class="t" style="width:450px;"><p>FictionLiveBench: my current fav third party long context benchmark. GPT-4.1 sits at 62.5%, while Gemini 2.5 Pro sits at 90.6%</p></td></tr></table></li><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"><span style="background-color:#e0e0e0;"><span style="color:rgb(255, 58, 58);font-size:0.6rem;">♥ 1k</span></span> Moonshot AI has announced Kimi-VL and Kimi-VL-Thinking, new open-source <b>Vision-Language models with reasoning capabilities</b>. The highlight of this release includes its multimodal reasoning and support for long context windows up to 128K tokens. It is now available via <a class="link" href="https://elink4f7.mail.bycloud.ai/ss/c/u001.CxDkkVpJsBdVoe83c_tBWno2StfHgkdXyV9QnRyXhktZQHs0gj7g4UeaEZ5_fNiyu5D8-eD-6Dbk4FiVdWKwyVca8MRbdKhGsrjJnU42z0yjUAVNGI3E-XUqDV8Nr3A-GWbZ7Vgf8UxvUNg1jAWFCZBUpbsjiscbaE5S5TXIabqTGphMT7elSOvF5SwZN4XK/4fo/GDpba6SsTIiMseR72g6Sfw/h2/h001.mzM8ryS2QSD9LYb3zUhKZYQKODfGtfCtsU_OrsFqUNc" target="_blank" rel="noopener noreferrer nofollow"><span>Hugging Face</span></a>, with their <a class="link" href="https://elink4f7.mail.bycloud.ai/ss/c/u001.VomAAYwkCjux8i_FMc4kJRdKc_ltU-p_qTm9zWFnRidcQt_iH1eLgzQqdwy8hOhdrQ1ma8HTZSLdiJoO1FmTSDooLkrvk3fqnHWuBeUuRf3E9AEZfa7GgjmJcHVVhM9h1Lc_cTC9v4ABBNJJhRWdqzK2sHe6kHTzVgODpEV9kfc/4fo/GDpba6SsTIiMseR72g6Sfw/h3/h001.8q8ZXqId9UVNzgKaiO76U2CfwQPBLYCyyiLbrF4jus4" target="_blank" rel="noopener noreferrer nofollow"><span>research paper available on GitHub</span></a>. </p><table role="none" border="0" cellspacing="0" cellpadding="0" style="margin:0 auto 0 auto;"><tr><td align="center" valign="top" style="width:480px;"><img src="https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/ae0aeaf0-ab3e-4d28-9bea-33b543dd2f9a/GoHLg9TaYAQWVZw.jpg?t=1744740637" alt="" height="auto" width="480" style="display:block;width:100%;" border="0"/></td></tr><tr><td align="center" valign="top" class="t" style="width:480px;"><p>Kimi-VL benchmarks</p></td></tr></table></li><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"><span style="background-color:#e0e0e0;"><span style="color:rgb(255, 58, 58);font-size:0.6rem;">♥ 1.1k</span></span> Google has shared <a class="link" href="https://elink4f7.mail.bycloud.ai/ss/c/u001.oB7zuO_W-X4Toa45C28ngyEFO5Gj_UpIFzWcxERDRtA6HJaTR3zWSZyRSG6EkPKQ6ZoHQpbSGDWtpaJyjo40yG5ClVCDAplXI9USJHOdUYRJXCAyccAFGw_KXuSLHpCTON8m6m39oJDp5xxzq23uDA/4fo/GDpba6SsTIiMseR72g6Sfw/h4/h001.BAPOXVLjTXpaVOM-zQHIsu9-Ri9ZLbAbcZQl5Vfu-pA" target="_blank" rel="noopener noreferrer nofollow"><span>DolphinGemma</span></a>, an AI model designed to analyze dolphin communication patterns. DolphinGemma utilizes the Open Gemma models and has been trained on The Dolphin Project’s acoustic database of wild Atlantic spotted dolphins. The model is capable of <b>processing complex sequences of dolphin sounds</b>, identifying patterns, and predicting likely subsequent sounds. </p><table role="none" border="0" cellspacing="0" cellpadding="0" style="margin:0 auto 0 auto;"><tr><td align="center" valign="top" style="width:480px;"><a href="https://elink4f7.mail.bycloud.ai/ss/c/u001.oB7zuO_W-X4Toa45C28ngyEFO5Gj_UpIFzWcxERDRtA6HJaTR3zWSZyRSG6EkPKQlIh_yKAjbFkcEsWj6_yNG0-SGuP_ledX401XyfhhY9mr4NGHU7d_BPqsRRKxfgTYrKkKsAoPBSUbwujATv8azw/4fo/GDpba6SsTIiMseR72g6Sfw/h5/h001.kFkN2tHSk5htYAn_a-BVFSKPv6EL4fTGjZSaADpOJYs" target="_blank" rel="noopener noreferrer nofollow" style="text-decoration:none;"><img src="https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/fdd8e161-d059-416c-9120-5a46ac8c64b2/Gof2B2UWUAERPzc.jpg?t=1744741134" alt="" height="auto" width="480" style="display:block;width:100%;" border="0"/></a></td></tr><tr><td align="center" valign="top" class="t" style="width:480px;"><p>dolphin sounds visualized</p></td></tr></table></li><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"><span style="background-color:#e0e0e0;"><span style="color:rgb(255, 58, 58);font-size:0.6rem;">♥ 1.1k</span></span> xAI has made <a class="link" href="https://elink4f7.mail.bycloud.ai/ss/c/u001.9ggl6Mt0xphuuMReR5gVpbqubpsoqd49s1VIoaEO_mc0854EtfXA_VCcR1wHoNKzMBXkJmjc5m934SXT3iTlZq9YX_LK0L9-vOfSqKthIEf2om9Kzs91ljLvZsgK62DYnfX2z29Rpy52SCRHj3T4fiaP1pcrAMm0xIQlQJ3OEGym381-yUArDEILkE4EOVRi/4fo/GDpba6SsTIiMseR72g6Sfw/h6/h001.TaLw00AG2AyaZB7e6nk5PBpi183oPRV__bBn1UnkJgU" target="_blank" rel="noopener noreferrer nofollow"><span>Grok-3 series available via the xAI API</span></a>, featuring text generation, image understanding, and image generation. These models support context windows up to <b>131,072 tokens</b>. </p><table role="none" border="0" cellspacing="0" cellpadding="0" style="margin:0 auto 0 auto;"><tr><td align="center" valign="top" style="width:420px;"><img src="https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/880173a7-9f70-40df-bd87-e9ad8bc996fd/GoIKc-1XEAA2T6P.jpg?t=1744741313" alt="" height="auto" width="420" style="display:block;width:100%;" border="0"/></td></tr><tr><td align="center" valign="top" class="t" style="width:420px;"><p>Grok-3 API pricing</p></td></tr></table></li></ol></div></td></tr><tr><td><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0" style=""><tr><td bgcolor="#222222" style="background-color:#222222;padding:0.0px 0.0px 0.0px 0.0px;"><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0"><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"></p></td></tr></table></td></tr></table></td></tr><tr><td><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0" style=""><tr><td bgcolor="transparent" style="background-color:transparent;border-color:#2C81E5;border-style:solid;border-width:5px;padding:0.0px 0.0px 0.0px 0.0px;"><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0"><tr><td class="dd" align="left" valign="top" style="color:#2A2A2A;font-weight:Bold;padding:0px 28px;text-align:left;"><h2 style="color:#2A2A2A;font-weight:Bold;"><span style="">Thunder Compute: The Cheapest cloud GPU</span></h2></td></tr><tr><td align="center" valign="top" style="padding-bottom:20px;padding-left:28px;padding-right:28px;padding-top:20px; " class="dd"><table role="none" border="0" cellspacing="0" cellpadding="0" style="margin:0 auto 0 auto;"><tr><td align="center" valign="top" style="width:600px;"><a href="https://elink4f7.mail.bycloud.ai/ss/c/u001.DUiN96-Eq7pUHzwEhy5j2-VOjK5bd-w_ZUTBDfQGUuEgB6UI_5qod7fjc4-Us-oxkJqYSG6PLW2d4B_BH0PLKcSi_nmeT5r04ad-pr0l3BMYuDIh9AsH9kBbFKw3PHSuI7UVOJtUOmCsjBwFuiwFcloYrh_S_ICMGh2u4GHPPYULOMpV-7tnwx0STiDPHXVuCojrNSTab0ri6fvS1ZQ1uw/4fo/GDpba6SsTIiMseR72g6Sfw/h7/h001.0GyZrBR67qZY_B_geTSskzABfJ3egGSLqLSX3zSWBtM" rel="noopener noreferrer nofollow" style="text-decoration:none;" target="_blank"><img src="https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/5fd53b29-8cbe-470a-9430-ccff7db9b921/mQrPqyfT.jpg?t=1744646625" alt="" height="auto" width="600" style="display:block;width:100%;" border="0"/></a></td></tr><tr><td align="center" valign="top" class="t" style="width:600px; padding: 4px 0px 4px 0px;"><p><span style="">A100 @ $0.57/hr!</span></p></td></tr></table></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"><span style=""><a class="link" href="https://elink4f7.mail.bycloud.ai/ss/c/u001.DUiN96-Eq7pUHzwEhy5j2-VOjK5bd-w_ZUTBDfQGUuEgB6UI_5qod7fjc4-Us-oxkJqYSG6PLW2d4B_BH0PLKcSi_nmeT5r04ad-pr0l3BMYuDIh9AsH9kBbFKw3PHSuhnPTzn7TTmzqu5qjGZlUquEShHIwX_7KxaukJmkVpn98S3p8mmnx5UYCBG77Yt85vqeCmto_Pop7FAOWa8k-Kw/4fo/GDpba6SsTIiMseR72g6Sfw/h8/h001.Us3_O8LL1rgdHO9zI9-Hm1D-Z5MuJHdETLnE-AZyK9g" target="_blank" rel="noopener noreferrer nofollow"><span>Thunder Compute</span></a></span><span style=""> is the cheapest way to get GPU cloud instances for AI, machine learning, or data science. You can get an </span><span style=""><b>A100 hosted</b></span><span style=""> in Google Cloud, in a </span><span style=""><b>US data center</b></span><span style="">, with best-in-class reliability and networking for</span><span style=""><b> $0.57/hr</b></span><span style="">, compared with $3.50/hr directly from Google.</span></p></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"><span style="">To make this possible, Thunder Compute invented virtualization software to network-attach GPUs. This increases the utilization of GPUs on the platform by 5x. Less downtime means lower prices for you.</span></p></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"><span style="">Thunder Compute uses a simple CLI to create and connect to instances. Just run </span><span style=""><code>tnr create</code></span><span style=""> and </span><span style=""><code>tnr connect [instance_id]</code></span><span style=""> to start.</span></p></td></tr><tr><td align="center" valign="top" style="padding-bottom:20px;padding-left:28px;padding-right:28px;padding-top:20px; " class="dd"><table role="none" border="0" cellspacing="0" cellpadding="0" style="margin:0 auto 0 auto;"><tr><td align="center" valign="top" style="width:600px;"><a href="https://elink4f7.mail.bycloud.ai/ss/c/u001.DUiN96-Eq7pUHzwEhy5j2-VOjK5bd-w_ZUTBDfQGUuEgB6UI_5qod7fjc4-Us-oxkJqYSG6PLW2d4B_BH0PLKcSi_nmeT5r04ad-pr0l3BMYuDIh9AsH9kBbFKw3PHSur_6-JhDKo4HGGzrRXjnAph6-YUlWwohfC4LTQdCUkHILcXE_JgyjLNFyT7OSPQ8Qa5Y1KO7lTkbAekL1xDMX0w/4fo/GDpba6SsTIiMseR72g6Sfw/h9/h001.LCsE71dwLJUpZ9C41exe1BFG1ZDI6dZOsNNvCC7IV0o" rel="noopener noreferrer nofollow" style="text-decoration:none;" target="_blank"><img src="https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/5d7604db-5bcf-4214-b7c8-44156d522aec/FUD1ZwR_.jpg?t=1744652666" alt="" height="auto" width="600" style="display:block;width:100%;" border="0"/></a></td></tr><tr><td align="center" valign="top" class="t" style="width:600px; padding: 4px 0px 4px 0px;"><p><span style="">it’s this easy to get started</span></p></td></tr></table></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"><span style="">You can use Thunder Compute through a simple CLI to create and connect to instances, or </span><span style=""><b>through their VSCode extension</b></span><span style=""> to develop directly on cloud GPUs in one click!</span></p></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"><span style=""><a class="link" href="https://elink4f7.mail.bycloud.ai/ss/c/u001.DUiN96-Eq7pUHzwEhy5j2-VOjK5bd-w_ZUTBDfQGUuEgB6UI_5qod7fjc4-Us-oxkJqYSG6PLW2d4B_BH0PLKcSi_nmeT5r04ad-pr0l3BMYuDIh9AsH9kBbFKw3PHSupMg9ym61SIje5SnzvFhqxYu_5v7WInzIk1rW6WR7_9Gk-IlfeTZUqcETR1BBCodCELf282VcIL-5opGzq9iJSQ/4fo/GDpba6SsTIiMseR72g6Sfw/h10/h001.uRKZyuETd154EznPHaPAe7DIYN0Pqc4OARMeu8MkHqE" target="_blank" rel="noopener noreferrer nofollow"><span>Create a </span></a></span><span style=""><a class="link" href="https://elink4f7.mail.bycloud.ai/ss/c/u001.DUiN96-Eq7pUHzwEhy5j2-VOjK5bd-w_ZUTBDfQGUuEgB6UI_5qod7fjc4-Us-oxkJqYSG6PLW2d4B_BH0PLKcSi_nmeT5r04ad-pr0l3BMYuDIh9AsH9kBbFKw3PHSuxuS5r9beYKExKg8eLGTSjgcYgv82H-asgcs10SV4rieFtPVQMFH_4yD-wlHZHwbnEQKFPj1kGexvlwfKpdy41g/4fo/GDpba6SsTIiMseR72g6Sfw/h11/h001.CZtYkGnxoqrr6mlSk3ANdNRE3MNWWjEkKbJQ1DuIl9A" target="_blank" rel="noopener noreferrer nofollow"><span>Thunder Compute</span></a></span><span style=""><a class="link" href="https://elink4f7.mail.bycloud.ai/ss/c/u001.DUiN96-Eq7pUHzwEhy5j2-VOjK5bd-w_ZUTBDfQGUuEgB6UI_5qod7fjc4-Us-oxkJqYSG6PLW2d4B_BH0PLKcSi_nmeT5r04ad-pr0l3BMYuDIh9AsH9kBbFKw3PHSuMsPACsL19KJDKV3pgkNA6CcYYbfZe7ksLjFY1QN_9l4lLKY3E8hc0SD2w1poWP96yvHFhj8wOLUPLhFgY_zm6A/4fo/GDpba6SsTIiMseR72g6Sfw/h12/h001.vXKCdXhdhnshGS1DKbBQItd7W2JLm_mYtbQdehgcjmI" target="_blank" rel="noopener noreferrer nofollow"><span> instance for </span></a></span><span style=""><b><a class="link" href="https://elink4f7.mail.bycloud.ai/ss/c/u001.DUiN96-Eq7pUHzwEhy5j2-VOjK5bd-w_ZUTBDfQGUuEgB6UI_5qod7fjc4-Us-oxkJqYSG6PLW2d4B_BH0PLKcSi_nmeT5r04ad-pr0l3BMYuDIh9AsH9kBbFKw3PHSu7LpK3NlVLHr3L3us-Ab2I8qbiHz0isCrR3XTcOSWQzxjhLcmdLK2h4rUn26AwMu9PCFq7Td5RbRByxQJkaGVUw/4fo/GDpba6SsTIiMseR72g6Sfw/h13/h001.4GD5j4FZlzODvp5ulk3Ch4nyqb44MBWoyXP5tCijG5M" target="_blank" rel="noopener noreferrer nofollow"><span>free</span></a></b></span><span style=""><b> with $20 per month of credit</b></span><span style="">.</span></p></td></tr><tr><td align="center" valign="top" style="padding-bottom:14px;padding-left:28px;padding-right:28px;padding-top:14px;text-align:center;width:100%;word-break:break-word;" class="dd"><table role="none" border="0" cellspacing="0" cellpadding="0" align="center" style="margin:14px auto 14px auto;"><tr><td align="center" valign="middle" height="44.75" style="height:44.75px;background-color:#2C81E5;border-color:#DFD150;border-radius:10px 10px 10px 10px;border-style:solid;border-width:0px 0px 0px 0px;color:#FFFFFF;font-family:'Open Sans','Segoe UI','Apple SD Gothic Neo','Lucida Grande','Lucida Sans Unicode',sans-serif;"><a href="https://elink4f7.mail.bycloud.ai/ss/c/u001.DUiN96-Eq7pUHzwEhy5j2-VOjK5bd-w_ZUTBDfQGUuEgB6UI_5qod7fjc4-Us-oxkJqYSG6PLW2d4B_BH0PLKcSi_nmeT5r04ad-pr0l3BMYuDIh9AsH9kBbFKw3PHSush9ax6Goal-NZUofHFCR_Xf9vHX3eSIpI1trTXNDzyzMhYwJ30WElgOKJBKsWBnuXLNGRXdOTFlFVnTvF2fjVA/4fo/GDpba6SsTIiMseR72g6Sfw/h14/h001.QCrBgZOw71KnWcuZcTneKw4yI48Xz7FFYcFkAD3mxw8" target="_blank" rel="noopener noreferrer nofollow" style="color:#FFFFFF;display:block;font-size:16px;font-size:16px;font-weight:normal;padding:0px 14px;padding:14px 14px 14px 14px;text-decoration:none;"> Check Out Thunder Compute </a></td></tr></table></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"><span style=""><a class="link" href="https://elink4f7.mail.bycloud.ai/ss/c/u001.tLfGW26lAwaS9gFg17HSoGymQ3NNPtd5dE5MV_8UgjIDFPVXngz8pvQBldSW42yhUe_Qiq6DgEPMEBuPL9yfRpXelTiuu2kS8pLFvsoem_XoZoy_n13sTKUhZIbl0VH6/4fo/GDpba6SsTIiMseR72g6Sfw/h15/h001.GYMG15y5d018zem_ZR3Y55FWpn7a4uQqRFPBJv-gaBE" target="_blank" rel="noopener noreferrer nofollow"><span>Advertise with The AI Timeline! </span></a></span></p></td></tr></table></td></tr></table></td></tr><tr><td><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0" style=""><tr><td bgcolor="#222222" style="background-color:#222222;padding:0.0px 0.0px 0.0px 0.0px;"><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0"><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"></p></td></tr></table></td></tr></table></td></tr><tr><td id="hogwild-inference-parallel-llm-gene" class="dd" align="left" valign="top" style="color:#2A2A2A;font-weight:Bold;padding:0px 28px;text-align:left;"><h2 style="color:#2A2A2A;font-weight:Bold;">Hogwild! Inference: Parallel LLM Generation via Concurrent Attention </h2></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"><span style=""><i>Rodionov et al. [Yandex, HSEUniversity, ITMO University, IST Austria]</i></span></p></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"><span style="background-color:#e0e0e0;"><span style="color:rgb(255, 58, 58);font-size:0.6rem;"> ♥ 317 </span></span><span style="color:rgb(44, 129, 229);font-size:0.6rem;"> </span><span style="background-color:#e0e0e0;"><span style="color:rgb(44, 129, 229);font-size:0.6rem;"> LLM Attention </span></span></p></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"> When humans tackle complex problems, we rarely work in isolation. We brainstorm, divide tasks, and adjust strategies dynamically. What if LLMs could do the same? Modern LLMs excel at tasks requiring long reasoning chains, but their sequential token-by-token generation limits efficiency. Previous attempts to parallelize LLM inference relied on predefined strategies: voting on answers, splitting problems into subtasks, or assigning specialized roles (e.g., "debugger" or "judge"). While these methods work for specific problems, they struggle when tasks don’t fit their structure. For instance, splitting a problem into subtasks fails if the initial plan is flawed, leaving workers stuck on irrelevant steps. </p></td></tr><tr><td align="center" valign="top" style="padding-bottom:20px;padding-left:28px;padding-right:28px;padding-top:20px; " class="dd"><table role="none" border="0" cellspacing="0" cellpadding="0" style="margin:0 auto 0 auto;"><tr><td align="center" valign="top" style="width:600px;"><img src="https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/f0c8af9c-8ac1-467c-9b31-dc0e8d3bff6d/image.png?t=1744730773" alt="" height="auto" width="600" style="display:block;width:100%;" border="0"/></td></tr></table></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"> This paper introduces <span style=""><i>Hogwild! Inference</i></span>, a method that allows multiple LLM instances to collaborate in parallel by sharing their "thoughts" in real time. This approach sidesteps rigid coordination frameworks, letting the models themselves decide how to work together. </p></td></tr><tr><td align="center" valign="top" style="padding:14px 32px 14px 32px;" class="dd"><a href="https://elink4f7.mail.bycloud.ai/ss/c/u001.VomAAYwkCjux8i_FMc4kJanfaI7V3KE8IM4ozPpTJ-iZ-_E4XBtjT8_VK6nAj9krjX8s94KNSEDV1QxwycZzIba1rznpZLb3sUGDHn4SrNZPi4n3tjw80AWFk41JDdOc/4fo/GDpba6SsTIiMseR72g6Sfw/h16/h001.eQZU1JBDhvjnIphGnnbQ-yylmDingrWAveIfLDMHkSc" style="text-decoration:none;" target="_blank"><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0" align="center" style="margin-bottom:12px;margin-top:12px;padding-left:12px;padding-right:12px;"><tr><td align="center" valign="top" class="o" style="padding:12px 12px 12px 12px;;background-color:#FFFFFF;border-color:#F1F1F1;border-radius:5px 5px 5px 5px;border-width:1px 1px 1px 1px;"><table role="none" border="0" cellspacing="0" cellpadding="0" align="right" width="100%"><tr><!--[if mso]><td width="0"><table cellpadding="0" cellspacing="0" border="0" role="presentation" style="display:none;"><tr><![endif]--><td class="embed-img" align="center" valign="top" style="width:100%;min-height:100px;vertical-align:middle;padding:0px 0px 12px 0px;"><a href="https://elink4f7.mail.bycloud.ai/ss/c/u001.VomAAYwkCjux8i_FMc4kJanfaI7V3KE8IM4ozPpTJ-iZ-_E4XBtjT8_VK6nAj9kryvLDzla1XvfjiJQC7uhklOyixMt09jY3mcpxqyH8r65cBifw9-N4RCpz3Ee85f9j/4fo/GDpba6SsTIiMseR72g6Sfw/h17/h001.F8NsA12wTnzQ4sE9vIxDZ6x5EGhTYVHXQPeg2GOT9Vo" style="text-decoration:none;" target="_blank"><img src="https://opengraph.githubassets.com/5eb966f0d63461234bd628fd6d2df40c0e0840590e54e12a095bc6e425789e07/eqimp/hogwild_llm" width="100%" style="display:block;"/></a></td><!--[if mso]></tr></table></td><![endif]--></tr><tr><td align="center" valign="top" class="cc"><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0" align="center"><tr><td align="left" valign="top" class="l"><p>GitHub - eqimp/hogwild_llm: Official PyTorch implementation for Hogwild! Inference: Parallel LLM Generation with a Concurrent Attention Cache</p></td></tr><tr><td align="left" valign="top" class="m"><p>Official PyTorch implementation for Hogwild! Inference: Parallel LLM Generation with a Concurrent Attention Cache - eqimp/hogwild_llm</p></td></tr></table></td></tr></table></td></tr></table></a></td></tr><tr><td id="how-hogwild-inference-works" class="dd" align="left" valign="top" style="color:#2A2A2A;font-weight:normal;padding:0px 28px;text-align:left;"><h3 style="color:#2A2A2A;font-weight:normal;"><span style="color:rgb(67, 67, 67);">How Hogwild! Inference Works</span></h3></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"> The Hogwild technique uses a shared <span style=""><i>Key-Value (KV) cache</i></span> that lets multiple LLM workers access each other’s intermediate outputs instantly. Instead of running isolated threads, these workers dynamically stitch their attention contexts together. To understand this better, let’s Imagine two assistants, Alice and Bob, solving a math problem: Alice might start by suggesting a task division, while Bob immediately notices an error in her approach and pivots. Their shared cache allows them to see and react to each other’s progress token-by-token. </p></td></tr><tr><td align="center" valign="top" style="padding-bottom:20px;padding-left:28px;padding-right:28px;padding-top:20px; " class="dd"><table role="none" border="0" cellspacing="0" cellpadding="0" style="margin:0 auto 0 auto;"><tr><td align="center" valign="top" style="width:600px;"><img src="https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/02e5aced-faea-4317-a05d-4b60e38d0eab/image.png?t=1744730808" alt="" height="auto" width="600" style="display:block;width:100%;" border="0"/></td></tr></table></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"> To make this practical, the method takes advantage of <span style=""><i>Rotary Position Embeddings (RoPE)</i></span>, which encode token positions as rotational angles in the attention mechanism. By rotating cached tokens to their new positions for each worker, Hogwild! avoids recomputing representations. The researchers tested three cache layouts for this: </p></td></tr><tr><td style="padding-bottom:12px;padding-left:50px;padding-right:40px;padding-top:12px;" class="ee"><div style="margin-left:0px;" class="edm_outlooklist"><ol start="1" style="list-style-type:decimal;margin:0px 0px;padding:0px 0px 0px 0px;"><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"><span style="font-weight:700;"><b>Contiguous</b></span>: Workers append tokens to private blocks, akin to collaborative document editing. </p></li><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"><span style="font-weight:700;"><b>Interleaved</b></span>: Workers share completed reasoning steps in a chat-like history. </p></li><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"><span style="font-weight:700;"><b>Combined</b></span>: A hybrid where workers see both real-time progress and shared history. </p></li></ol></div></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"> The system prompt encourages collaboration by periodically asking workers to check for redundant work (e.g., “Wait, am I doing redundant work?”). Surprisingly, models like QwQ-32B and DeepSeek-R1 adapt naturally to this setup and they often redistribute tasks or revise plans without explicit training. </p></td></tr><tr><td id="performance-and-tradeoffs" class="dd" align="left" valign="top" style="color:#2A2A2A;font-weight:normal;padding:0px 28px;text-align:left;"><h3 style="color:#2A2A2A;font-weight:normal;"><span style="color:rgb(67, 67, 67);">Performance and Trade-offs</span></h3></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"> The researchers tested Hogwild technique on a number of synthetic and complex reasoning tasks (e.g., GSM8k and LIMO datasets): </p></td></tr><tr><td style="padding-bottom:12px;padding-left:50px;padding-right:40px;padding-top:12px;" class="ee"><div style="margin-left:0px;" class="edm_outlooklist"><ul style="font-weight:normal;list-style-type:disc;margin-bottom:12px !important;margin-top:12px !important;padding:0px 0px 0px 0px;"><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"> With a token budget of 4096, Hogwild!’s <span style=""><i>combined layout</i></span> <span style="font-weight:700;"><b>solved 68.2%</b></span> of LIMO tasks, outperforming independent workers (48.4%) and single-threaded baselines (52.3%). </p></li><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"> Smaller budgets (1024 tokens) favor the <span style=""><i>contiguous layout</i></span>, where immediate synchronization helps workers coordinate faster. </p></li><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"> At higher budgets (8192 tokens), the <span style=""><i>interleaved layout</i></span> catches up, as step-wise synchronization reduces noise from overlapping token streams. </p></li></ul></div></td></tr><tr><td align="center" valign="top" style="padding-bottom:20px;padding-left:28px;padding-right:28px;padding-top:20px; " class="dd"><table role="none" border="0" cellspacing="0" cellpadding="0" style="margin:0 auto 0 auto;"><tr><td align="center" valign="top" style="width:600px;"><img src="https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/fccba800-cf5e-4ca0-8c58-d869ca7bae06/image.png?t=1744730850" alt="" height="auto" width="600" style="display:block;width:100%;" border="0"/></td></tr></table></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"> However, there are caveats. Coordination overhead can hurt performance with too many workers: four workers underperformed a single worker on synthetic tasks due to excessive time spent negotiating roles. </p></td></tr><tr><td align="center" valign="top" style="padding-bottom:14px;padding-left:28px;padding-right:28px;padding-top:14px;text-align:center;width:100%;word-break:break-word;" class="dd"><table role="none" border="0" cellspacing="0" cellpadding="0" align="center" style="margin:14px auto 14px auto;"><tr><td align="center" valign="middle" height="44.75" style="height:44.75px;background-color:#2C81E5;border-color:#DFD150;border-radius:10px 10px 10px 10px;border-style:solid;border-width:0px 0px 0px 0px;color:#FFFFFF;font-family:'Open Sans','Segoe UI','Apple SD Gothic Neo','Lucida Grande','Lucida Sans Unicode',sans-serif;"><a href="https://elink4f7.mail.bycloud.ai/ss/c/u001.DUiN96-Eq7pUHzwEhy5j28yjf9KIXZdsXoh1WlHvvKn3mguBqxVZQOK7O4qjQh6vsoWObVGeEDOIt9iMSQBkTCUiqECx_JD7JDsogmuDgPS_WK2xLFmyb9cf3kpn_Lsk/4fo/GDpba6SsTIiMseR72g6Sfw/h18/h001.QXWVvCMckSnhXN0TwL06-AXDG64KNnTaTprS9cvnReE" target="_blank" rel="noopener noreferrer nofollow" style="color:#FFFFFF;display:block;font-size:16px;font-size:16px;font-weight:normal;padding:0px 14px;padding:14px 14px 14px 14px;text-decoration:none;"> Read Full Paper </a></td></tr></table></td></tr><tr><td><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0" style=""><tr><td bgcolor="#222222" style="background-color:#222222;padding:0.0px 0.0px 0.0px 0.0px;"><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0"><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"></p></td></tr></table></td></tr></table></td></tr><tr><td id="one-minute-video-generation-with-te" class="dd" align="left" valign="top" style="color:#2A2A2A;font-weight:Bold;padding:0px 28px;text-align:left;"><h2 style="color:#2A2A2A;font-weight:Bold;">One-Minute Video Generation with Test-Time Training</h2></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"><span style=""><i>Dalal et al. [NVIDIA, Stanford University, UCSD, UC Berkeley, UT Austin]</i></span></p></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"><span style="background-color:#e0e0e0;"><span style="color:rgb(255, 58, 58);font-size:0.6rem;"> ♥ 5.7k </span></span><span style="color:rgb(44, 129, 229);font-size:0.6rem;"> </span><span style="background-color:#e0e0e0;"><span style="color:rgb(44, 129, 229);font-size:0.6rem;"> Video Generation </span></span><span style="color:rgb(44, 129, 229);font-size:0.6rem;"> </span></p></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"> Video generation models are getting better day by day, but telling coherent, multi-scene stories longer than 30 seconds is still pretty challenging. This is because in Transformers’, self-attention scales poorly with context length, while modern RNN alternatives like Mamba lack the expressive power to handle dynamic motion and complex scene transitions. </p></td></tr><tr><td align="center" valign="top" style="padding-bottom:20px;padding-left:28px;padding-right:28px;padding-top:20px; " class="dd"><table role="none" border="0" cellspacing="0" cellpadding="0" style="margin:0 auto 0 auto;"><tr><td align="center" valign="top" style="width:600px;"><img src="https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/76776121-f60b-4381-866c-0dd79d6ceed4/image.png?t=1744730951" alt="" height="auto" width="600" style="display:block;width:100%;" border="0"/></td></tr></table></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"> A new approach with Test-Time Training (TTT) layers uses a hybrid approach that treats hidden states as trainable neural networks. By updating these states through gradient descent during inference, the model dynamically adapts to retain critical story elements across scenes. </p></td></tr><tr><td align="center" valign="top" class="dd"><a href="https://elink4f7.mail.bycloud.ai/ss/c/u001.tLfGW26lAwaS9gFg17HSoHadyiR_3hx5dNWvnYT4zroGjlkq7shctCRl2fq5vPPDPiJRlwCYCQzt83uzYkYxFvjQ6IVZTN-0_cnuD4ZJyfZDo_oApzh6hefiftoLMuJxbOKW9RyJTC_WVT_e5uNHNg/4fo/GDpba6SsTIiMseR72g6Sfw/h19/h001.NcMMG-Y-35UQPCWVTa1Sup4FZrfJY-bRPnMKA81HYl4" style="text-decoration:none;" target="_blank"><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0" align="center" style="margin-bottom:12px;margin-top:12px;padding-left:12px;padding-right:12px;"><tr><td align="center" valign="top" class="o" style="padding:12px 12px 12px 12px;;background-color:#FFFFFF;border-color:#F1F1F1;border-radius:5px 5px 5px 5px;border-width:1px 1px 1px 1px;"><table role="none" border="0" cellspacing="0" cellpadding="0" align="right" width="100%"><tr><!--[if mso]><td width="0"><table cellpadding="0" cellspacing="0" border="0" role="presentation" style="display:none;"><tr><![endif]--><td class="embed-img mob-stack" align="center" valign="top" style="width:35%;min-height:100px;vertical-align:middle;display:none;"><a href="https://elink4f7.mail.bycloud.ai/ss/c/u001.tLfGW26lAwaS9gFg17HSoHadyiR_3hx5dNWvnYT4zroGjlkq7shctCRl2fq5vPPDdNgKLJ-RtcObPZPbt7lwrThIxDrEjA-GeCdAI47th4EGJodVWvZW3liD6tQHPTT4s3LjMdUIVlNnHMRLXwxILg/4fo/GDpba6SsTIiMseR72g6Sfw/h20/h001.UzzF7Gqg-pyId1SmiD_27aPtVV_031V-nXhbwsl26HM" style="text-decoration:none;" target="_blank"><img src="" width="100%" style="display:block;"/></a></td><!--[if mso]></tr></table></td><![endif]--></tr><tr><td align="center" valign="top" class="cc"><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0" align="center"><tr><td align="left" valign="top" class="l"><p>One-Minute Video Generation with Test-Time Training</p></td></tr><tr><td align="left" valign="top" class="m"><p>A new approach using Test-Time Training (TTT) layers to generate coherent, minute-long videos from text.</p></td></tr></table></td><!--[if mso]><td width="0"><table cellpadding="0" cellspacing="0" border="0" role="presentation" style="display:none;"><tr><![endif]--><td class="mob-hide" align="center" valign="top" style="width:35%;min-height:100px;padding:0px 0px 0px 12px;vertical-align:middle;"><img src="" width="100%" style="display:block;"/></td><!--[if mso]></tr></table></td><![endif]--></tr></table></td></tr></table></a></td></tr><tr><td id="how-neural-hidden-states-in-test-ti" class="dd" align="left" valign="top" style="color:#2A2A2A;font-weight:normal;padding:0px 28px;text-align:left;"><h3 style="color:#2A2A2A;font-weight:normal;"><span style="color:rgb(67, 67, 67);">How Neural Hidden States in Test-Time Training Remember Better</span></h3></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"> TTT layers reimagine the hidden state as a two-layer MLP that evolves with each frame. Unlike static matrices in Mamba or DeltaNet, this MLP trains on-the-fly using a self-supervised task: reconstructing corrupted versions of input frames. For each token in the sequence, the layer: </p></td></tr><tr><td style="padding-bottom:12px;padding-left:50px;padding-right:40px;padding-top:12px;" class="ee"><div style="margin-left:0px;" class="edm_outlooklist"><ol start="1" style="list-style-type:decimal;margin:0px 0px;padding:0px 0px 0px 0px;"><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"><span style="font-weight:700;"><b>Corrupts</b></span> the input (e.g., masking parts of a frame), </p></li><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"><span style="font-weight:700;"><b>Updates</b></span> the hidden MLP by minimizing reconstruction loss, </p></li><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"><span style="font-weight:700;"><b>Predicts</b></span> the original input using the refined MLP. </p></li></ol></div></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"> This process embeds a learning loop directly into the forward pass. The hidden state creates a model that actively adapts to fill in gaps, preserving details like character movements and scene layouts. </p></td></tr><tr><td align="center" valign="top" style="padding-bottom:20px;padding-left:28px;padding-right:28px;padding-top:20px; " class="dd"><table role="none" border="0" cellspacing="0" cellpadding="0" style="margin:0 auto 0 auto;"><tr><td align="center" valign="top" style="width:600px;"><img src="https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/80cdb3ac-c9ac-45f1-aef9-6f307b33cd44/image.png?t=1744731004" alt="" height="auto" width="600" style="display:block;width:100%;" border="0"/></td></tr></table></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"> To integrate TTT into existing architectures, researchers added gated connections to a pre-trained 5B-parameter Diffusion Transformer. Self-attention layers handle local 3-second segments, while TTT layers stitch these segments globally. A bidirectional processing trick allows the model to learn from both past and future context without violating causality, crucial for maintaining continuity across scene cuts. </p></td></tr><tr><td align="center" valign="top" style="padding-bottom:20px;padding-left:28px;padding-right:28px;padding-top:20px; " class="dd"><table role="none" border="0" cellspacing="0" cellpadding="0" style="margin:0 auto 0 auto;"><tr><td align="center" valign="top" style="width:600px;"><img src="https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/9958d083-7280-4e02-8293-faff233b3095/image.png?t=1744731044" alt="" height="auto" width="600" style="display:block;width:100%;" border="0"/></td></tr></table></td></tr><tr><td id="evaluating-test-time-training-appro" class="dd" align="left" valign="top" style="color:#2A2A2A;font-weight:normal;padding:0px 28px;text-align:left;"><h3 style="color:#2A2A2A;font-weight:normal;"><span style="color:rgb(67, 67, 67);">Evaluating Test-Time Training Approach</span></h3></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"> In human evaluations, TTT-based videos outperformed Mamba 2 and sliding-window attention by 34 Elo points, which is comparable to GPT-4 over GPT-3.5. During testing, the researchers observed that: </p></td></tr><tr><td style="padding-bottom:12px;padding-left:50px;padding-right:40px;padding-top:12px;" class="ee"><div style="margin-left:0px;" class="edm_outlooklist"><ul style="font-weight:normal;list-style-type:disc;margin-bottom:12px !important;margin-top:12px !important;padding:0px 0px 0px 0px;"><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"><span style="font-weight:700;"><b>Temporal consistency</b></span>: Characters maintained appearance across scene changes. </p></li><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"><span style="font-weight:700;"><b>Motion naturalness</b></span>: Dynamic actions (e.g., chases) flowed smoothly between segments. </p></li><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"><span style="font-weight:700;"><b>Story adherence</b></span>: Multi-step plots (like Jerry stealing a pie through coordinated tricks) stayed on track. </p></li></ul></div></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"> However, there are still a few artifacts in the videos. Additionally, TTT layers are not as efficient and they add 2.5× inference latency compared to Mamba, though they’re far cheaper than full self-attention. </p></td></tr><tr><td align="center" valign="top" style="padding-bottom:20px;padding-left:28px;padding-right:28px;padding-top:20px; " class="dd"><table role="none" border="0" cellspacing="0" cellpadding="0" style="margin:0 auto 0 auto;"><tr><td align="center" valign="top" style="width:600px;"><img src="https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/cb81b2bc-8e0b-4619-bab3-be18867a5b68/image.png?t=1744731075" alt="" height="auto" width="600" style="display:block;width:100%;" border="0"/></td></tr><tr><td align="center" valign="top" class="t" style="width:600px; padding: 4px 0px 4px 0px;"><p>Human evaluation results for one-minute videos.</p></td></tr></table></td></tr><tr><td align="center" valign="top" style="padding-bottom:14px;padding-left:28px;padding-right:28px;padding-top:14px;text-align:center;width:100%;word-break:break-word;" class="dd"><table role="none" border="0" cellspacing="0" cellpadding="0" align="center" style="margin:14px auto 14px auto;"><tr><td align="center" valign="middle" height="44.75" style="height:44.75px;background-color:#2C81E5;border-color:#DFD150;border-radius:10px 10px 10px 10px;border-style:solid;border-width:0px 0px 0px 0px;color:#FFFFFF;font-family:'Open Sans','Segoe UI','Apple SD Gothic Neo','Lucida Grande','Lucida Sans Unicode',sans-serif;"><a href="https://elink4f7.mail.bycloud.ai/ss/c/u001.DUiN96-Eq7pUHzwEhy5j28yjf9KIXZdsXoh1WlHvvKnd7LjdkvnYPMdscYaSsVD3IN0q-3vkOgmHBR-ptQWLcOBtkxqAk8sycU_VXkVrM2E8JxdDA748B0WGP5dv1RLL/4fo/GDpba6SsTIiMseR72g6Sfw/h21/h001.10an2BJhWXU4v0LojHFYcrXFYFoinD142oNTsVLOPDg" target="_blank" rel="noopener noreferrer nofollow" style="color:#FFFFFF;display:block;font-size:16px;font-size:16px;font-weight:normal;padding:0px 14px;padding:14px 14px 14px 14px;text-decoration:none;"> Read Full Paper </a></td></tr></table></td></tr><tr><td><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0" style=""><tr><td bgcolor="#222222" style="background-color:#222222;padding:0.0px 0.0px 0.0px 0.0px;"><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0"><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"></p></td></tr></table></td></tr></table></td></tr><tr><td id="gaussian-mixture-flow-matching-mode" class="dd" align="left" valign="top" style="color:#2A2A2A;font-weight:Bold;padding:0px 28px;text-align:left;"><h2 style="color:#2A2A2A;font-weight:Bold;">Gaussian Mixture Flow Matching Models</h2></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"><span style=""><i>Chen et al. [Stanford University, Adobe Research, Hillbot]</i></span></p></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"><span style="background-color:#e0e0e0;"><span style="color:rgb(255, 58, 58);font-size:0.6rem;"> ♥ 301 </span></span><span style="color:rgb(44, 129, 229);font-size:0.6rem;"> </span><span style="background-color:#e0e0e0;"><span style="color:rgb(44, 129, 229);font-size:0.6rem;"> Image Generation </span></span><span style="color:rgb(44, 129, 229);font-size:0.6rem;"> </span><span style="background-color:#e0e0e0;"><span style="color:rgb(44, 129, 229);font-size:0.6rem;"> bycloud’s pick </span></span><span style="color:rgb(44, 129, 229);font-size:0.6rem;"> </span></p></td></tr><tr><td id="introduction-to-gaussian-mixture-fl" class="dd" align="left" valign="top" style="color:#2A2A2A;font-weight:normal;padding:0px 28px;text-align:left;"><h3 style="color:#2A2A2A;font-weight:normal;"><span style="color:rgb(67, 67, 67);">Introduction to Gaussian Mixture Flow Matching (GMFlow)</span></h3></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"> Generative models are getting better but even state-of-the-art methods face stubborn challenges: generating high-quality samples in just a few steps often leads to artifacts, and popular guidance techniques like classifier-free guidance (CFG) tend to oversaturate colors. </p></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"> Traditional diffusion and flow matching models simplify the denoising process by assuming the distribution of "flow velocity" (the direction and speed at which noise transitions to data) follows a single Gaussian. This works well when taking tiny, incremental steps during sampling. But when you try to generate images in just a handful of steps, that approximation breaks down. Large step sizes introduce errors, leading to blurry or distorted outputs. </p></td></tr><tr><td align="center" valign="top" style="padding-bottom:20px;padding-left:28px;padding-right:28px;padding-top:20px; " class="dd"><table role="none" border="0" cellspacing="0" cellpadding="0" style="margin:0 auto 0 auto;"><tr><td align="center" valign="top" style="width:600px;"><img src="https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/c5173d16-99a2-4d9a-bf6e-cfdfb8528f80/example_results.png?t=1744731295" alt="" height="auto" width="600" style="display:block;width:100%;" border="0"/></td></tr></table></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"><span style="font-weight:700;"><b>Gaussian Mixture Flow Matching (GMFlow)</b></span> addresses these limitations by replacing the single-Gaussian assumption with a <span style="font-weight:700;"><b>Gaussian mixture model (GMM)</b></span>. Instead of predicting a single flow velocity, GMFlow estimates a multi-modal distribution of velocities. This allows the model to capture complex, overlapping pathways from noise to data which enables higher-quality generation with fewer steps while avoiding guidance-induced artifacts. </p></td></tr><tr><td align="center" valign="top" style="padding:14px 32px 14px 32px;" class="dd"><a href="https://elink4f7.mail.bycloud.ai/ss/c/u001.VomAAYwkCjux8i_FMc4kJd9ntkKacl64gN-nVD8_a_xtNOUJmvd30l0iVmdedqvR5uyokRdD_Ljq2X22sAfMW_96gxG_TkbuvfmA0sMWpaEpiqXXEGEQhgmfZlCexAni/4fo/GDpba6SsTIiMseR72g6Sfw/h22/h001.oVxxSwFWDYoPclbGL4wlA4wDGhH05MSHOztCfH1QKoY" style="text-decoration:none;" target="_blank"><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0" align="center" style="margin-bottom:12px;margin-top:12px;padding-left:12px;padding-right:12px;"><tr><td align="center" valign="top" class="o" style="padding:12px 12px 12px 12px;;background-color:#FFFFFF;border-color:#F1F1F1;border-radius:5px 5px 5px 5px;border-width:1px 1px 1px 1px;"><table role="none" border="0" cellspacing="0" cellpadding="0" align="right" width="100%"><tr><!--[if mso]><td width="0"><table cellpadding="0" cellspacing="0" border="0" role="presentation" style="display:none;"><tr><![endif]--><td class="embed-img" align="center" valign="top" style="width:100%;min-height:100px;vertical-align:middle;padding:0px 0px 12px 0px;"><a href="https://elink4f7.mail.bycloud.ai/ss/c/u001.VomAAYwkCjux8i_FMc4kJd9ntkKacl64gN-nVD8_a_xtNOUJmvd30l0iVmdedqvRqhzMZd4Ws12S1jI9kcUEkz95ELjMAApT4Xmc9_LPTFawZBmAipaYWNjMuScRzkNL/4fo/GDpba6SsTIiMseR72g6Sfw/h23/h001.2Zxx7wMcMmj8MQzVtAIJNXagqHD4-RkVbBVTlY8CcBY" style="text-decoration:none;" target="_blank"><img src="https://opengraph.githubassets.com/65de3a9a564bd5dc07f7dace42c20608a452f35481b12e637e9cc7e892f3c99e/Lakonik/GMFlow" width="100%" style="display:block;"/></a></td><!--[if mso]></tr></table></td><![endif]--></tr><tr><td align="center" valign="top" class="cc"><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0" align="center"><tr><td align="left" valign="top" class="l"><p>GitHub - Lakonik/GMFlow: Gaussian Mixture Flow Matching Models (GMFlow)</p></td></tr><tr><td align="left" valign="top" class="m"><p>Gaussian Mixture Flow Matching Models (GMFlow). Contribute to Lakonik/GMFlow development by creating an account on GitHub.</p></td></tr><tr><td align="left" valign="bottom" class="n" style="vertical-align:bottom;padding-top:12px;"><p style="word-break:break-word;">github.com/Lakonik/GMFlow</p></td></tr></table></td></tr></table></td></tr></table></a></td></tr><tr><td id="how-gm-flow-works-mixtures-guidance" class="dd" align="left" valign="top" style="color:#2A2A2A;font-weight:normal;padding:0px 28px;text-align:left;"><h3 style="color:#2A2A2A;font-weight:normal;"><span style="color:rgb(67, 67, 67);">How GMFlow Works: Mixtures, Guidance, and Analytic Solvers</span></h3></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"><span style="color:rgb(67, 67, 67);">There are three parts in GMFlow framework: </span></p></td></tr><tr><td id="modeling-multi-modal-velocity-distr" class="dd" align="left" valign="top" style="color:#2A2A2A;font-weight:normal;padding:0px 28px;text-align:left;"><h4 style="color:#2A2A2A;font-weight:normal;"><span style="color:rgb(102, 102, 102);">Modeling Multi-Modal Velocity Distributions</span></h4></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"> GMFlow predicts a mixture of Gaussians to represent the possible flow velocities at each denoising step. For every noisy input, the model outputs: </p></td></tr><tr><td style="padding-bottom:12px;padding-left:50px;padding-right:40px;padding-top:12px;" class="ee"><div style="margin-left:0px;" class="edm_outlooklist"><ul style="font-weight:normal;list-style-type:disc;margin-bottom:12px !important;margin-top:12px !important;padding:0px 0px 0px 0px;"><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"><span style="font-weight:700;"><b>Means</b></span> (directions) for each Gaussian component. </p></li><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"><span style="font-weight:700;"><b>Weights</b></span> (probabilities) for selecting among components. </p></li><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"> A shared <span style="font-weight:700;"><b>variance</b></span> (spread) across all components. </p></li></ul></div></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"> Training uses a KL divergence loss to align the predicted mixture with the true velocity distribution. This generalizes traditional flow matching, which uses an L2 loss to regress a single mean. By capturing multi-modality, GMFlow better approximates the true denoising dynamics, even when large steps are taken. </p></td></tr><tr><td align="center" valign="top" style="padding-bottom:20px;padding-left:28px;padding-right:28px;padding-top:20px; " class="dd"><table role="none" border="0" cellspacing="0" cellpadding="0" style="margin:0 auto 0 auto;"><tr><td align="center" valign="top" style="width:600px;"><img src="https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/844c83cb-277b-4b0c-94e0-28cb1dc794ed/image.png?t=1744731356" alt="" height="auto" width="600" style="display:block;width:100%;" border="0"/></td></tr></table></td></tr><tr><td id="probabilistic-guidance" class="dd" align="left" valign="top" style="color:#2A2A2A;font-weight:normal;padding:0px 28px;text-align:left;"><h4 style="color:#2A2A2A;font-weight:normal;"><span style="color:rgb(102, 102, 102);">Probabilistic Guidance</span></h4></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"> Classifier-free guidance (CFG) amplifies conditional signals by extrapolating between conditional and unconditional predictions. However, this extrapolation often pushes samples outside the training data distribution, causing oversaturation. GMFlow’s <span style="font-weight:700;"><b>probabilistic guidance</b></span> avoids this by reweighting the Gaussian mixture components instead of extrapolating. </p></td></tr><tr><td style="padding-bottom:12px;padding-left:50px;padding-right:40px;padding-top:12px;" class="ee"><div style="margin-left:0px;" class="edm_outlooklist"><ol start="1" style="list-style-type:decimal;margin:0px 0px;padding:0px 0px 0px 0px;"><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"> The model estimates both conditional (e.g., "a cat") and unconditional ("an image") velocity distributions as GMMs. </p></li><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"> Guidance strengthens the conditional signal by <span style=""><i>reweighting</i></span> the mixture probabilities toward components that align with the condition. </p></li><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"> Unlike CFG, this keeps samples within the data distribution, preventing oversaturation while improving alignment. </p></li></ol></div></td></tr><tr><td align="center" valign="top" style="padding-bottom:20px;padding-left:28px;padding-right:28px;padding-top:20px; " class="dd"><table role="none" border="0" cellspacing="0" cellpadding="0" style="margin:0 auto 0 auto;"><tr><td align="center" valign="top" style="width:600px;"><img src="https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/a4642983-d801-49e6-812c-11c3cb140aa8/image.png?t=1744731453" alt="" height="auto" width="600" style="display:block;width:100%;" border="0"/></td></tr></table></td></tr><tr><td id="sampling-with-analytic-precision" class="dd" align="left" valign="top" style="color:#2A2A2A;font-weight:normal;padding:0px 28px;text-align:left;"><h4 style="color:#2A2A2A;font-weight:normal;"><span style="color:rgb(102, 102, 102);">Sampling with Analytic Precision</span></h4></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"> GMFlow introduces specialized solvers (GM-SDE and GM-ODE) that leverage the analytic properties of Gaussian mixtures. When predicting the next denoising step, these solvers: </p></td></tr><tr><td style="padding-bottom:12px;padding-left:50px;padding-right:40px;padding-top:12px;" class="ee"><div style="margin-left:0px;" class="edm_outlooklist"><ol start="1" style="list-style-type:decimal;margin:0px 0px;padding:0px 0px 0px 0px;"><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"> Compute the exact transition distribution by combining the predicted mixture components. </p></li><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"> Use closed-form solutions to integrate velocity fields, reducing discretization errors. </p></li></ol></div></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"> This allows GMFlow to take larger steps without sacrificing accuracy. For example, in a 2D toy experiment, GMFlow reconstructs a checkerboard pattern in just four steps, while traditional methods require 16+ steps to avoid severe artifacts. </p></td></tr><tr><td align="center" valign="top" style="padding-bottom:20px;padding-left:28px;padding-right:28px;padding-top:20px; " class="dd"><table role="none" border="0" cellspacing="0" cellpadding="0" style="margin:0 auto 0 auto;"><tr><td align="center" valign="top" style="width:600px;"><img src="https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/6de1629d-c2cf-4128-9be4-b502da15d519/image.png?t=1744731408" alt="" height="auto" width="600" style="display:block;width:100%;" border="0"/></td></tr><tr><td align="center" valign="top" class="t" style="width:600px; padding: 4px 0px 4px 0px;"><p> Comparison among vanilla flow models with different solvers and GMFlow.</p></td></tr></table></td></tr><tr><td id="benchmark-results-for-gm-flow" class="dd" align="left" valign="top" style="color:#2A2A2A;font-weight:normal;padding:0px 28px;text-align:left;"><h3 style="color:#2A2A2A;font-weight:normal;"><span style="color:rgb(67, 67, 67);">Benchmark Results for GMFlow</span></h3></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"> GMFlow challenges the long-held assumption that single-Gaussian dynamics are sufficient for diffusion and flow models. By embracing multi-modality, it opens the door to faster, and better sampling. GMFlow was evaluated on two benchmarks: a synthetic 2D dataset and ImageNet 256×256 generation. </p></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"><span style="font-weight:700;"><b>2D Checkerboard Analysis</b></span></p></td></tr><tr><td style="padding-bottom:12px;padding-left:50px;padding-right:40px;padding-top:12px;" class="ee"><div style="margin-left:0px;" class="edm_outlooklist"><ul style="font-weight:normal;list-style-type:disc;margin-bottom:12px !important;margin-top:12px !important;padding:0px 0px 0px 0px;"><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"> With just four steps, GMFlow nearly perfectly reconstructs the checkerboard, while baseline methods (DDPM, DPM++) show blurred or fragmented patterns. </p></li><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"> Increasing the number of Gaussians (K) improves sample fidelity. K=64 achieves near-perfect results, while K=1 (equivalent to standard flow matching) struggles. </p></li></ul></div></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"><span style="font-weight:700;"><b>ImageNet 256×256</b></span></p></td></tr><tr><td style="padding-bottom:12px;padding-left:50px;padding-right:40px;padding-top:12px;" class="ee"><div style="margin-left:0px;" class="edm_outlooklist"><ul style="font-weight:normal;list-style-type:disc;margin-bottom:12px !important;margin-top:12px !important;padding:0px 0px 0px 0px;"><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"> GMFlow achieves a Precision score of <span style="font-weight:700;"><b>0.942</b></span> with only six sampling steps, outperforming flow matching baselines by a wide margin. At 32 steps, it reaches a state-of-the-art Precision of <span style="font-weight:700;"><b>0.950</b></span>. </p></li><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"> Probabilistic guidance cuts saturation levels by 30% compared to CFG, aligning outputs closer to natural image statistics. </p></li><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"> Despite its complexity, GMFlow adds minimal computational overhead, just 0.005 seconds per step on an A100 GPU. </p></li></ul></div></td></tr><tr><td align="center" valign="top" style="padding-bottom:20px;padding-left:28px;padding-right:28px;padding-top:20px; " class="dd"><table role="none" border="0" cellspacing="0" cellpadding="0" style="margin:0 auto 0 auto;"><tr><td align="center" valign="top" style="width:600px;"><img src="https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/uploads/asset/file/419e091b-1238-4290-8988-1b8d6c4513f3/image.png?t=1744731502" alt="" height="auto" width="600" style="display:block;width:100%;" border="0"/></td></tr><tr><td align="center" valign="top" class="t" style="width:600px; padding: 4px 0px 4px 0px;"><p>ImageNet evaluation results at best Precision</p></td></tr></table></td></tr><tr><td class="dd" align="left" style="padding:0px 28px;text-align:left;word-break:break-word;"><p style="line-height:24px;"><span style="font-weight:700;"><b>Limitations and Trade-offs</b></span></p></td></tr><tr><td style="padding-bottom:12px;padding-left:50px;padding-right:40px;padding-top:12px;" class="ee"><div style="margin-left:0px;" class="edm_outlooklist"><ul style="font-weight:normal;list-style-type:disc;margin-bottom:12px !important;margin-top:12px !important;padding:0px 0px 0px 0px;"><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"><span style="font-weight:700;"><b>Pixel-Wise Factorization</b></span>: GMFlow models each pixel independently, which simplifies training but ignores spatial correlations. The authors propose spectral sampling to inject spatial coherence, though this remains an area for improvement. </p></li><li class="listItem ultext"><p style="line-height:24px;padding:0px;text-align:left;word-break:break-word;"><span style="font-weight:700;"><b>Component Sensitivity</b></span>: Performance plateaus at around K=8 components for images, suggesting diminishing returns beyond a certain complexity. </p></li></ul></div></td></tr><tr><td align="center" valign="top" style="padding-bottom:14px;padding-left:28px;padding-right:28px;padding-top:14px;text-align:center;width:100%;word-break:break-word;" class="dd"><table role="none" border="0" cellspacing="0" cellpadding="0" align="center" style="margin:14px auto 14px auto;"><tr><td align="center" valign="middle" height="44.75" style="height:44.75px;background-color:#2C81E5;border-color:#DFD150;border-radius:10px 10px 10px 10px;border-style:solid;border-width:0px 0px 0px 0px;color:#FFFFFF;font-family:'Open Sans','Segoe UI','Apple SD Gothic Neo','Lucida Grande','Lucida Sans Unicode',sans-serif;"><a href="https://elink4f7.mail.bycloud.ai/ss/c/u001.DUiN96-Eq7pUHzwEhy5j28yjf9KIXZdsXoh1WlHvvKn8jCoJOxZ0_tcEp3Vtnv5tLOIyMYTWSLHW-ecKmgVyNDjRPGdUlS0XKbXv9ql6gmbrHdQB3Brat9pbOYwwPRSc/4fo/GDpba6SsTIiMseR72g6Sfw/h24/h001.Pf54BJy92eell1VJLyUsesrc6fSW5p7_AACpCRU-2Os" target="_blank" rel="noopener noreferrer nofollow" style="color:#FFFFFF;display:block;font-size:16px;font-size:16px;font-weight:normal;padding:0px 14px;padding:14px 14px 14px 14px;text-decoration:none;"> Read Full Paper </a></td></tr></table></td></tr><tr><td class="dd" style="padding: 20px;"><table width="100%" cellpadding="0" cellspacing="0" role="none" style="max-width:520px;margin:0 auto;"><tr><td class="q" style="padding:16px 16px 6px 16px;"><a href="https://elink4f7.mail.bycloud.ai/ss/c/u001.tLfGW26lAwaS9gFg17HSoDDFT6eh5Nsg0xYVQj-h6I3o9m2k79_qw4izMYhmcI36gKbtdWUp-f16fD_Jz5UbNEGYTA4GO8gEJgA1Z8py5mReGS1iYlMyDYlPpW1aFMI89EWSHMJFSNDS6OHgA6253GdtP2p0xolRRMJwsbynuOI/4fo/GDpba6SsTIiMseR72g6Sfw/h25/h001.qT7YC5XWXQHCBXnClo_JuQouSS696-xg88n_3onSbuU" style="text-decoration:none !important;"><table width="100%" cellpadding="0" cellspacing="0" border="0" role="none"><tr><td width="100%" style="padding: 0 0 14px 0;text-decoration:none;width:100%;"><table width="100%" cellpadding="0" cellspacing="0" border="0" role="none"><tr><td width="36" style="width:36px;"><img src="https://pbs.twimg.com/profile_images/1698572487909400576/BvncwnrP_normal.jpg" alt="tw profile: The AI Timeline" style="display:block;width:36px;height:36px;border-radius:50%;border:0;"/></td><td width="400" style="padding:0 0 0 8px;text-decoration:none;"><span style="display:block;font-size:14px;color:#1c2022;font-weight:700;"> The AI Timeline </span><span style="display:block;color:#697882;font-size:14px;"> @TheAITimeline </span></td><td width="24" align="right" style="vertical-align:text-top;"><img width="24" height="24" loading="lazy" alt="tw" style="border:0;" src="https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/static_assets/x_logo.png"/></td></tr></table></td></tr><tr></tr><tr><td style="word-break:break-word;"><p>🚨This week's top AI/ML research papers:</p><p>- Scaling Laws for Native Multimodal Models <br>- Quantization Hurts Reasoning? <br>- The AI Scientist -v2 <br>- Parallel LLM Generation via Concurrent Attention <br>- Gaussian Mixture Flow Matching Models <br>- VAPO <br>- Are Reasoning Models losing Critical</p></td></tr><tr><td style="padding:12px 0 0 0;"></td></tr><tr><td align="center" style="padding:8px 0 0 0;width:480px;"><img src="https://pbs.twimg.com/media/God3X14XcAAN0WH.jpg" width="480" height="auto" style="display:block;border:1px solid #E1E8ED;border-radius:5px;width:100%;max-width:480px;height:auto;"/></td></tr><tr><td height="8" style="line-height:1px;font-size:1px;height:8px;"> </td></tr><tr><td align="left" valign="top" class="s"><p>4:10 AM • Apr 14, 2025</p></td></tr><tr><td height="10" style="line-height: 1px; font-size: 1px; height: 10px;"> </td></tr><tr><td height="1" bgcolor="#e1e8ed" style="line-height:0px;font-size:0px;height:1px;"></td></tr><tr><td height="10" style="line-height:1px;font-size:1px;height:10px;"> </td></tr><tr><td align="left" valign="top" class="s"><p><b style="color:#1C2022">395</b> Likes <b style="color:#1C2022">60</b> Retweets </p></td></tr><tr><td align="left" valign="top" class="s"><div align="center" style="text-align:center;margin-top:4px;margin-bottom:4px;padding:8px;border:1px solid #ccd6dd;border-radius:9999px;color:#1B95E0"><b>6 Replies</b></div></td></tr></table></a></td></tr></table></td></tr><tr><td class="dd" align="center" valign="top" style="padding:20px;"><a href="https://elink4f7.mail.bycloud.ai/ss/c/u001.amatuKKICSickUKplYJXmPuLmBHhwTO6tQ9Brgw0IfeR2OqMa_L7QOLYZWsp25A2XaE8qB38vFwt21FtUJf02CJfwRXYbPFvvdRaKp2BubUS43s2rLpXo340imswKquh/4fo/GDpba6SsTIiMseR72g6Sfw/h26/h001.3LqEDNvWUGtZ3UsdrR0na3M5wY_lmXFp0i7ULxez_gU" style="text-decoration:none;"><table align="center" width="100%" cellpadding="0" cellspacing="0" border="0" role="none" style="max-width:520px;margin:0 auto;"><tr><td class="p" width="100%" style="padding:2px;border:none;"><table width="100%" cellpadding="0" cellspacing="0" border="0" role="none"><tr><td align="center" valign="top" style="width:100%;"><div style="max-height:0;position:relative;opacity:0.999;width:100%;mso-hide:all;"><div style="display:inline-block;width:100%;padding-top:25%;"><img width="20%" height="auto" loading="lazy" alt="" style="border:0;" src="https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/static_assets/youtube_play_icon.png"/></div></div><a href="https://elink4f7.mail.bycloud.ai/ss/c/u001.amatuKKICSickUKplYJXmPuLmBHhwTO6tQ9Brgw0Ifc0JwIrmZyXXHowRMtZ5STblpxnxPRkdKMvqab98N7tF5KN1en3t5j8Pcdvj6krqkGHzZzyhvV-nFlViStXx9yl/4fo/GDpba6SsTIiMseR72g6Sfw/h27/h001.zPzhzLYW18hB6vr2fP-e2vWqccKV19uEa4C1mAYZiDo" style="text-decoration:none;"><img src="https://i.ytimg.com/vi/SRgER6-7yYQ/maxresdefault.jpg" width="480" height="auto" loading="lazy" alt="YouTube video by bycloud" style="display:block;height:auto;border:0;outline:none;text-decoration:none;background-color:#000000;width:100%;"/></a></td></tr><tr><td><p style="font-size:12px;font-weight:500;font-style:italic;font-family:Helvetica, Calibri, sans-serif;color: #686a6d; padding-top:0 !important;padding-bottom:6px !important; padding-left:4px !important;"> OpenAI’s amazing naming scheme strikes again (feat. GPT-4.1) </p></td></tr></table></td></tr></table></a></td></tr></table></td></tr></table></td></tr><tr><td align="center" valign="top"><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0" align="center"><tr><td><tr><td class="b" align="center" valign="top" bgcolor="#2a2a2a" style="padding:0px 0px 0px 0px;border-style:solid;border-width: 0px 0px 0px 0px;border-color: #2a2a2a;border-bottom-left-radius:10px;border-bottom-right-radius:10px;"><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0" align="center"><tr><td align="center" valign="top" bgcolor="#73ddff" style="padding:12px"><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0" align="center"><tr><td><span style="padding-left:1px;"></span></td><td align="center" valign="middle" width="75" style="width:75px;"><a href="https://elink4f7.mail.bycloud.ai/ss/c/u001.1muhFWIqieRYpaJ-FbWSCQqcWoV4NNHHr5SkP9THApWuHAAlWLQxI3Q_IqFmt_DcyAxeC8jDApCnHmMSBGpBb5sgtimvBYgxRX-Rp7s0F3LjCHoSwdhr83OBqRFhJ1y_/4fo/GDpba6SsTIiMseR72g6Sfw/h28/h001.Tf1m39dgFceHKSEWOgWdpJ6h9pYp02e6KHY4161i0pE" style="text-decoration:none;"><img width="22" height="22" alt="tw" border="0" style="display:block;max-width:22px;color:Dark" src="https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/static_assets/x_dark.png"/></a></td><td align="center" valign="middle" width="75" style="width:75px;"><a href="https://elink4f7.mail.bycloud.ai/ss/c/u001.amatuKKICSickUKplYJXmBoQnQ9VXnB2zTxBG4HeHBgjMqVxpoXRdj01cjwyoVlHgiebEOgBvwHtevoVpsSvpn3Q1di2ml6sb3cBM-X6IStQbj_zQSVGWJ8AAmPw2en2/4fo/GDpba6SsTIiMseR72g6Sfw/h29/h001.oWlnlUqoYH49qkYjjU1otKnoY1SBvRNvQYPoqcHoiR4" style="text-decoration:none;"><img width="22" height="16" alt="yt" border="0" style="display:block;max-width:22px;color:Dark" src="https://media.beehiiv.com/cdn-cgi/image/fit=scale-down,format=auto,onerror=redirect,quality=80/static_assets/youtube_dark.png"/></a></td><td><span style="padding-left:1px;"></span></td></tr></table></td></tr><tr><td height="10" style="line-height:1px;font-size:1px;height:10px;"> </td></tr><tr><td class="w" align="center" valign="top" style="padding:15px 15px 15px 15px;"><table role="none" width="100%" border="0" cellspacing="0" cellpadding="0" align="center"><tr><td align="center" valign="top"><p style="font-family:'Verdana',Geneva,sans-serif;color:#FFFFFF!important;"> Update your email preferences or unsubscribe <a class="link" href="https://elink4f7.mail.bycloud.ai/ss/c/u001.c6q0w4g5sodbtO4I1B_pxWc4htTObwdorovK0nFHVH-4pUdVE0ELYH5DsNemk732SjNwhPNJ25r0O8B5vYifsBhEpz-DJgyVFmavJPa0OyKRRnvw4o7XGyvIv7PRofnmCXKjJbBin19BPSH-my1pOkHAcs7SAq-vBcm3Evt_-KwpMqKinmJvIS37SOtjTI4VLTTTDMBj3tYqoqGr8g9QeHwwn5ITeyMkcKTJmIMOnYnVVm4_kBCRmJucU9iOnIbU-OAiFpaVEmcrVePqaU6-1mv6N99W0uAlaTU23gwnKII5_tng118Zt6ddM1jcqym66gUoH3q2Y9XGOL5LF0BNs9sEkAXuXG-1XwkGz9LA6r-rldWEvoZy6P3G8Ygo5fYFaR5B1_O6V5RXZXjURThBVarmR0rXdlX3cXJINJb6MDyBuUa6I3XOF1pgBD9W2LmzdmbO4Ptt0TPsYhDP9j9AeZD1-K3kKqHAvyc7gSMHILSdAnpZhoqLFg0xi5ZMErsvgG9yRsAP4iKRstSZlM1ZPWFggCmV9Nh2Z3OqOHdrklQSPe9Mw_rk6TuxkbDzk7RHbd_4v4jBmKuxS9-YxBbg_3TS5_wIDTeUr1mXGCI_dmVePCEsyisTfYF5I-7Y1uCYd1_75pirVB8vMEJC-KXwqmw9OahdpPEg0dhMhYxkNZmMYJ0yCq4Q028hPWJbKnhnN9Xzw4H4uRQ-9tS-pPvcxgfY5BmURNdymGT0VOmjY-lw35YDs8grRl9oIqu_3gJyjoewgf_N8JBAEgNQ27JwlO5rhJQ98fgjej-KcjbBWfjbE6By5m3w7WaYkEHYqjw1dWh84X25M83Ostsg4UODtw/4fo/GDpba6SsTIiMseR72g6Sfw/h30/h001.ZvnjL5oZT1ASpWYQmQyvuYFQBmKmvBmZB8-DoZt40hU" style="text-decoration:underline;text-decoration-color:#FFFFFF!important;color:#FFFFFF!important;"> here</a></p><p class="copyright" style="font-family:'Verdana',Geneva,sans-serif;color:#FFFFFF!important;"> © 2025 bycloudai </p><p style="font-family:'Verdana',Geneva,sans-serif;color:#FFFFFF!important;"> 228 Park Ave S, #29976, New York, New York 10003, United States </p></td></tr><tr style="display: table-row !important;"><td align="center" valign="top" style="padding-top:20px;" style="display:table-cell !important;"><table role="none" border="0" cellspacing="0" cellpadding="0" align="center" style="display:table !important;"><tr style="display:table-row !important;"><td class="u" align="center" valign="middle" height="32" style="height:32px;display:table-cell !important; max-height: 32px !important;margin:0px !important; background-color: #ffffff !important;"><a style="line-height:32px !important;text-decoration:none;display:block !important;" href="https://elink4f7.mail.bycloud.ai/ss/c/u001.DUiN96-Eq7pUHzwEhy5j28olDWFpV5DDKfdk_OdOKOju4V0lTqdlVQQu9BgTRIDAGkf12lk2kDJ5tDC_ldRxoPzd-ZUJbEu4ELUIvDpismJwmDcSYwR-DKEouXzJCff_NR00bckdGwsbqK-CGoir2sSgX1Wl40MalkX_yf_KFJQ3dCKNzSHXeIJWKEGyI-p0z5FtvH1GlXE6rwIaj4ljOxSDB_kT6SDNLmBcXP7sYIO-pY69YEiGOGF6oQbF-9qI/4fo/GDpba6SsTIiMseR72g6Sfw/h31/h001.7qs31cTWFSGEBoyildwzGnOrkH25ym6sxGvkaiUSNMA"><img src="https://media.beehiiv.com/output-onlinepngtools.png" width="16" alt="beehiiv logo" style="display:inline-block !important;max-width:16px !important; vertical-align:-3px !important;width: 16px !important;" border="0"/><span style="padding-left:11px !important;display: inline-block !important;">Powered by beehiiv</span></a></td></tr></table></td></tr><tr><td align="left" valign="top" height="2" style="height:2px;"><a href='https://elink4f7.mail.bycloud.ai/ss/c/u001.CxDkkVpJsBdVoe83c_tBWsHIaP4XNp0WgUYqLvHcKk_3uqk_KIkz4ddLinhFbud6JuxLFdSUhYnR7b1NSsmbtzXNGNblnEEMKUtkCAjkn8Y/4fo/GDpba6SsTIiMseR72g6Sfw/h32/h001.a3YBxj_QeGmEP2L996Qv2N-kVCqTsnpOzZ9E5CwBSzg' style="color: #2a2a2a !important; cursor: default; font-size: 1px; text-decoration: none;"> Terms of Service </a></td></tr></table></td></tr></table></td></tr></td></tr></table></td></tr></table></td></tr></table></td></tr></table></div></body></html>