--- title: Why can’t LLMs just LEARN the context window? | daily.dev description: A research paper called 'End-to-End Test-Time Training for Long Context' proposes storing context information directly into a transformer's MLP weights via... canonical: https://daily.dev/posts/why-can-t-llms-just-learn-the-context-window--h4w2jvmkl twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Why can’t LLMs just LEARN the context window? | daily.dev og:description: A research paper called 'End-to-End Test-Time Training for Long Context' proposes storing context information directly into a transformer's MLP weights via... og:url: https://daily.dev/posts/why-can-t-llms-just-learn-the-context-window--h4w2jvmkl og:image: https://api.daily.dev/og/posts/H4w2JvMkL.png og:image:alt: Why can’t LLMs just LEARN the context window? og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Why can’t LLMs just LEARN the context window? **[bycloud](https://daily.dev/sources/bycloud)** · 12 min read · 7 upvotes · 1 comments ## Summary A research paper called 'End-to-End Test-Time Training for Long Context' proposes storing context information directly into a transformer's MLP weights via gradient updates at inference time, rather than relying on full attention or KV cache. The approach combines sliding window attention for local context with test-time weight updates for long-range context, making it computationally feasible through batching. Experiments show it tracks full attention performance closely at 32K–128K token contexts and maintains a consistent loss advantage. However, precise needle-in-a-haystack retrieval remains a weakness compared to full attention, since information is compressed into weights without indexing. The method is presented as a promising new direction toward effectively infinite context windows through continual learning. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Similar posts on daily.dev - [Reimagining LLM Memory: Using Context as Training Data Unlocks Models That Learn at Test-Time](https://daily.dev/posts/reimagining-llm-memory-using-context-as-training-data-unlocks-models-that-learn-at-test-time-w4lvqtrk7) · NVIDIA Developer · 1 upvotes · 0 comments - [How LLMs Handle Infinite Context With Finite Memory](https://daily.dev/posts/how-llms-handle-infinite-context-with-finite-memory-4eiofc662) · Towards Data Science · 2 upvotes · 0 comments - [Context window in AI: why every token is a budget decision](https://daily.dev/posts/context-window-in-ai-why-every-token-is-a-budget-decision-reskx1u2b) · Redis · 0 upvotes · 0 comments --- Tags: [#llm](https://daily.dev/tags/llm) [View this post on daily.dev](https://daily.dev/posts/why-can-t-llms-just-learn-the-context-window--h4w2jvmkl) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Why can’t LLMs just LEARN the context window?","url":"https://daily.dev/posts/why-can-t-llms-just-learn-the-context-window--h4w2jvmkl","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/why-can-t-llms-just-learn-the-context-window--h4w2jvmkl"},"datePublished":"2026-03-30T20:27:34.882Z","dateModified":"2026-03-30T20:27:57.224Z","description":"A research paper called 'End-to-End Test-Time Training for Long Context' proposes storing context information directly into a transformer's MLP weights via...","image":"https://i.ytimg.com/vi/P9uNy71YukQ/sddefault.jpg","thumbnailUrl":"https://i.ytimg.com/vi/P9uNy71YukQ/sddefault.jpg","isAccessibleForFree":true,"articleSection":"bycloud","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"bycloud","logo":"https://media.daily.dev/image/upload/s--UUgAibgb--/f_auto/v1712315122/logos/bycloud","url":"https://daily.dev/sources/bycloud"},"commentCount":1,"discussionUrl":"https://daily.dev/posts/why-can-t-llms-just-learn-the-context-window--h4w2jvmkl","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":7},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":1}],"keywords":"llm","timeRequired":"PT12M","video":{"@type":"VideoObject","name":"Why can’t LLMs just LEARN the context window?","description":"A research paper called 'End-to-End Test-Time Training for Long Context' proposes storing context information directly into a transformer's MLP weights via...","thumbnailUrl":"https://i.ytimg.com/vi/P9uNy71YukQ/sddefault.jpg","uploadDate":"2026-03-30T20:27:34.882Z","duration":"PT12M","url":"https://api.daily.dev/r/H4w2JvMkL","embedUrl":"https://www.youtube.com/embed/P9uNy71YukQ"}} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"bycloud","item":"https://daily.dev/sources/bycloud"},{"@type":"ListItem","position":3,"name":"Why can’t LLMs just LEARN the context window?"}]} ```