Did Codex Reset
GitHub

DAIR.AI highlights Meta’s RankEvolve, which pairs Claude Code and Codex

DAIR.AI

DAIR.AI describes RankEvolve, from a Meta paper on reliable auto-research agents, as a system that enforces each research phase and gate through a compiled protocol. It runs Claude Code and Codex as separate nodes that review and repair each other’s changes, aimed at silent experiment bugs such as leaked evaluation data or a disconnected gradient.

At a matched budget, combining the two products raises execution accuracy from 45.8% for the best single product to 62.5%. Over twelve iterations on the open-source HSTU recommender, RankEvolve improved NDCG@10 on MovieLens-20M by 4.48% over the published result. The write-up links to the RankEvolve paper on DAIR.AI Academy.

#rankevolve#meta#claude-code#codex#auto-research

Meta Superintelligence Labs paper describes a Sharpening Tax in RL post-training

Across 42 base and post-trained pairs, base models with a light harness often solve more agentic tasks at large sample counts than their RL post-trained versions, which still win on pass@1. The authors’ PTGS method sets each prompt’s sampling temperature from estimated difficulty during RL, paying a smaller tax and also raising pass@1.