Grace

gracekind.net

04 October 2026 07:46

This paper provides a (much needed!) explanation of how the attention mechanism relates to Birch's "persisting interlocutor illusion", and a good summary of the LLM individuation problem overall:

Max Puelma Touzel

mptouzel.bsky.social

04 October 2026 07:46

The most recent Kambhampati paper on the illusion of reasoning of reasoning traces (out of distribution at least) lnkd.in/p/gZfFX9HU

Marc Lanctot

sharky6000.bsky.social

03 October 2026 22:52

Any #commodore64 gamers out there? I have got just the thing to awaken the October spirits: Castles of Dr. Creep Longplay with a spoooooky soundtrack spready across the ages! 😅 youtu.be/ZHw1-GRM2Jg?...

Nathan Lambert

natolambert.bsky.social

02 October 2026 17:28

Today we're unveiling Trillium Labs, a new non-profit to foster the open science of frontier AI. We're building open post-training recipes and will expand into open infra to study RSI, reward-hacking, multi-agent systems, and whatever comes next. blog.trilliumlabs.org/p/introducin...

Max Kleiman-Weiner

maxkw.bsky.social

02 October 2026 16:30

Everyone is excited about recursive *self* improvement (RSI) but human progress is social and driven by cumulative culture. Individuals stand on the shoulders of past giants. Our new paper tests whether LLM swarms benefit from an alternative RSI: recursive *social* improvement.