Create accountLogin

Recent searches

No recent searches

Search options

Only available when logged in.
veganism.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
Veganism Social is a welcoming space on the internet for vegans to connect and engage with the broader decentralized social media community.

Administered by:

Server stats:

292
active users

veganism.social: AboutStatusPrivacy policy

Mastodon: AboutGet the appKeyboard shortcutsView source codev4.3.0-alpha.5+vegan

#contextwindow

1 post1 participant0 posts today
Jim Groth<p>Just published a small experiment called ThinContext:<br>A simple way to reduce context bloat in LLMs by summarizing previous outputs instead of feeding everything back verbatim.</p><p>message → full reply<br>context → short summary<br>Only context gets reused.</p><p>It’s lossy, but functional.<br>Maybe a small step toward more meaningful memory in AI systems.</p><p>Repo here if you’re curious, or want to break it: <a href="https://github.com/VanDerGroot/ThinContext" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">github.com/VanDerGroot/ThinCon</span><span class="invisible">text</span></a></p><p>Critique very welcome.</p><p><a href="https://mastodon.social/tags/AI" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AI</span></a> <a href="https://mastodon.social/tags/LLM" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>LLM</span></a> <a href="https://mastodon.social/tags/ContextWindow" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>ContextWindow</span></a> <a href="https://mastodon.social/tags/PromptEngineering" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>PromptEngineering</span></a> <a href="https://mastodon.social/tags/OpenSource" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>OpenSource</span></a></p>
N-gated Hacker News<p>🚀🐪 <a href="https://mastodon.social/tags/Meta" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Meta</span></a> just dropped Llama 4, a "groundbreaking" 788GB masterpiece that's guaranteed to gather digital dust on your hard drive. 🤔💾 With its "affordable" <a href="https://mastodon.social/tags/10M" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>10M</span></a> context window and the intelligence that rivals a walnut, Llama 4 is the ultimate test of our collective patience! 😂🔍<br><a href="https://www.llama.com/llama4/" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="">llama.com/llama4/</span><span class="invisible"></span></a> <a href="https://mastodon.social/tags/Llama4" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Llama4</span></a> <a href="https://mastodon.social/tags/groundbreaking" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>groundbreaking</span></a> <a href="https://mastodon.social/tags/digitaldust" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>digitaldust</span></a> <a href="https://mastodon.social/tags/AItechnology" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AItechnology</span></a> <a href="https://mastodon.social/tags/contextwindow" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>contextwindow</span></a> <a href="https://mastodon.social/tags/HackerNews" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>HackerNews</span></a> <a href="https://mastodon.social/tags/ngated" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>ngated</span></a></p>
Nicole Hennig<p>Hands on with Gemini 2.5 Pro: why it might be the most useful reasoning model yet <a href="https://venturebeat.com/ai/beyond-benchmarks-gemini-2-5-pro-is-probably-the-best-reasoning-model-yet/" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">venturebeat.com/ai/beyond-benc</span><span class="invisible">hmarks-gemini-2-5-pro-is-probably-the-best-reasoning-model-yet/</span></a> <a href="https://techhub.social/tags/AI" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AI</span></a> <a href="https://techhub.social/tags/reasoning" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>reasoning</span></a> <a href="https://techhub.social/tags/ContextWindow" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>ContextWindow</span></a></p>
Erik Jonker<p>Fascinating, must-read blog by Steven Johnson (from NotebookLM), about the importance and impact of the ever growing context window in LLMs.<br><a href="https://thelongcontext.com/" rel="nofollow noopener noreferrer" translate="no" target="_blank"><span class="invisible">https://</span><span class="">thelongcontext.com/</span><span class="invisible"></span></a><br><a href="https://mastodon.social/tags/AI" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AI</span></a> <a href="https://mastodon.social/tags/LLM" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>LLM</span></a> <a href="https://mastodon.social/tags/contextwindow" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>contextwindow</span></a> <a href="https://mastodon.social/tags/stevenjohnson" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>stevenjohnson</span></a> <a href="https://mastodon.social/tags/NotebookLM" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>NotebookLM</span></a></p>
VegansExploreLive feeds
About