💯

Wanna understand the sources of uncertainty in LLMs when performing in-context learning 🤔? 🚀 We introduce a variational uncertainty decomposition framework for in-context learning without explicitly sampling from the latent parameter posterior. 📄 Paper:

I was recently asked during an interview why LLMs were not deterministic even when temperature is 0. This very carefully answers it! My interviewer’s answer was different though, that there likely is non determinism in MOE routing for load balancing

ChatGPT doesn't understand why I keep beating it at rock paper scissors😅

This will be an oral! 🎤 See you at #EMNLP25

🚨New Meta Superintelligence Labs Paper🚨 What do we do when we don’t have reference answers for RL? What if annotations are too expensive or unknown? Compute as Teacher (CaT🐈) turns inference compute into a post-training supervision signal. CaT improves up to 30% even on

VUD’s gonna be at #NeurIPS2025 🎉🥳 Special thanks to my labmates that made this collaboration especially enjoyable!

Delighted to announce that our work, ShiQ: Bringing back Bellman to LLMs has been accepted to NeurIPS 2025 🥳🥳🎉🎉🎊🎊 Details of the paper in the thread below 👇👇

Available by Dec 2025: Oliver Ratmann and I are looking for 1 PhD student on AI for pathogen deep-sequence analytics. Scholarship includes UK home fee + stipend. Contact my work email if interested. RT 🙏 docs.google.com/document/u/0/d…

Arjun and Vagrant will be presenting our work at Session 3 (11AM 🕚 poster session) at Conference on Language Modeling. If you're interested in LLM evaluation practices, probability vs. generation-based evals, harms and misgendering, please go say hi 👋! Link to paper 📜: arxiv.org/abs/2504.17075

Post-training methods like RLHF improve LLM quality but often collapse diversity. Check out DQO, a training objective using DPPs that directly optimizes for semantic diversity and quality.

Thanks for having me! Fantastic to see such innovative research happening at Imperial College London and inspiring to meet so many brilliant students shaping the future of AI research!

🚨 New Paper: The Art of Scaling Reinforcement Learning Compute for LLMs 🚨 We burnt a lot of GPU-hours to provide the community with the first open, large-scale systematic study on RL scaling for LLMs. x.com/Devvrit_Khatri…