A curated archive of frontier intelligence, operator-grade guides, and strategic analysis.
New research introduces Foresighted Policy Optimization (FPO) to prevent alignment collapse in iterative RLHF, addressing how LLMs exploit reward model blind spots.
Read the briefing
FREIA, a new unsupervised reinforcement learning algorithm, improves LLM reasoning by adaptively balancing consensus and exploration, outperforming baselines in mathematical...
New arXiv research models how In-Context Learning (ICL) in transformers generalizes out-of-distribution (OOD). It finds that task vectors in a...
