Thrilled to share our new preprint on Reinforcement Learning for Reverse Engineering (RLRE) 🚀 We demonstrate that human preferences can be reverse engineered effectively by pipelining LLMs to optimise upstream preambles via reinforcement learning 🧵⬇️

I'm excited to share the tech report for our @cohere.com @cohereforai.bsky.social Command A and Command R7B models. We highlight our novel approach to model training including self-refinement algorithms and model merging techniques at scale. Read more below! ⬇️

I really enjoyed my MLST chat with Tim @neuripsconf.bsky.social about the research we've been doing on reasoning, robustness and human feedback. If you have an hour to spare and are interested in AI robustness, it may be worth a listen 🎧 Check it out at youtu.be/DL7qwmWWk88?...

Check out @lisaalaz.bsky.social's internship work with us @cohere.com questioning the rationale behind rationales 🔥

Lisa Alazraki@lisaalaz.bsky.social · last yr.

Do LLMs need rationales for learning from mistakes? 🤔 When LLMs learn from previous incorrect answers, they typically observe corrective feedback in the form of rationales explaining each mistake. In our new preprint, we find these rationales do not help, in fact they hurt performance! 🧵

Super excited to see PRISM recognised as a #NeurIPS2024 best paper. This was an incredible large-scale effort by @hannahrosekirk.bsky.social and fantastic collaborators. If you're interested in human feedback, check it out, there are 100+ pages of detailed insights! 🔥

NeurIPS Conference@neuripsconf.bsky.social · 2y ago

Announcing the NeurIPS 2024 Best Paper Awards:

Our paper PRISM alignment won a best paper award at #neurips2024! All credits to @hannahrosekirk.bsky.social A.Whitefield, P.Röttger, A.M.Bean, K.Margatina, R.Mosquera-Gomez, J.Ciro, @maxbartolo.bsky.social H.He, B.Vidgen, S.Hale Catch Hannah tomorrow at neurips.cc/virtual/2024/poster/97804

Excited to reveal Genie 2, our most capable foundation world model that, given a single prompt image, can generate an endless variety of action-controllable, playable 3D worlds. Fantastic cross-team effort by the Open-Endedness Team and many other teams at Google DeepMind! 🧞

Jack Parker-Holder@jparkerholder.bsky.social · 2y ago

Introducing 🧞Genie 2 🧞 - our most capable large-scale foundation world model, which can generate a diverse array of consistent worlds, playable for up to a minute. We believe Genie 2 could unlock the next wave of capabilities for embodied agents 🧠.

Looking forward to @neuripsconf.bsky.social #NeurIPS #NeurIPS2024 in Vancouver next week! ❄️ Reach out (or pop by the @cohere.com booth) if you want to chat about human feedback, robustness and reasoning, prompt optimisation, adversarial data, glitch tokens, evaluation, or anything else!

an advertisement for vancouver in british columbia canada

ALT: an advertisement for vancouver in british columbia canada

media.tenor.com

Sparks of multi-hop reasoning ✨

Sohee Yang@soheeyang.bsky.social · 2y ago

🚨 New Paper 🚨 Can LLMs perform latent multi-hop reasoning without exploiting shortcuts? We find the answer is yes – they can recall and compose facts not seen together in training or guessing the answer, but success greatly depends on the type of the bridge entity (80% for country, 6% for year)! 1/N

Fun to see Douwe's Dynabench plot continue to inspire new groundbreaking benchmarking work!

Tim Rocktäschel@handle.invalid · 2y ago

Excited to announce "BALROG: a Benchmark for Agentic LLM and VLM Reasoning On Games" led b UCL DARK's @dpaglieri.bsky.social! Douwe Kiela plot below is maybe the scariest for AI progress — LLM benchmarks are saturating at an accelerating rate. BALROG to the rescue. This will keep us busy for years.

🚨 LLMs can learn to reason from procedural knowledge in pretraining data! 🚨 I particularly enjoy research where the evidence contradicts our initial hypothesis. If you're interested in LLM reasoning, check out the 60+ pages of in-depth work at arxiv.org/abs/2411.12580

Laura@lauraruis.bsky.social · 2y ago

How do LLMs learn to reason from data? Are they ~retrieving the answers from parametric knowledge🦜? In our new preprint, we look at the pretraining data and find evidence against this: Procedural knowledge in pretraining drives LLM reasoning ⚙️🔢 🧵⬇️

We launched Judge Arena with @huggingface.bsky.social @clefourrier.bsky.social - a platform that lets you easily compare models as judges side-by-side and vote for the best evaluation Check out the live leaderboard and start voting now 🤗