Alana Renda @ICLR26 🇧🇷

Humans augment learning from experience and demonstrations with learning from general instructions, rules, and descriptions of behaviour. PBB enables LLMs to do the same, unlocking a highly sample efficient form of learning. Excited to share this work in Brazil! 🇧🇷

🚨 New paper up on how LLMs reason under uncertainty! 🎲 Many real world uses of LLMs are characterized by the unknown—not only are the models prompted with partial information, but often even humans don't know the "right answer" to the questions asked. Yet most LLM evals focus on problems with clearly defined success criteria. There’s a gap in our understanding of how models perform in this setting. We investigate.... 🔎

🚨New Paper!🚨 We trained reasoning LLMs to reason about what they don't know. o1-style reasoning training improves accuracy but produces overconfident models that hallucinate more. Meet RLCR: a simple RL method that trains LLMs to reason and reflect on their uncertainty -- improving both accuracy ✅ and calibration 🎯. [1/N]