RL-XAR results: 89% and 95% human win rates on papers and stories, judged across model families
jaseweston · x · 2026-09-28
- Meta's RL-XAR thread (4/5) details training: Qwen3.5-27B trained with RL on learned rubrics scored by a cross-family Qwen3.8-2.4T-A95B judge, with final results verified by GPT5-6.
- Strong gains across scientific paper sections, story continuations and Wikipedia pages, largest on the first two.
- Small human evals show 89% (papers) and 95% (stories) win rates for RL-XAR; training fixes over-scoping and section focus in papers, removes clichéd writing in stories.
Related event: Meta's RL-XAR Uses Expert-Aligned Rubrics to Fix AI Slop(4 posts)→
More from Research
- Google Research open-sources PaperOrchestra, a multi-agent framework for writing AI papers — google-research · 2026-11-16
- ai-rulez author shows how a single success score hides MCP skill-discovery failure modes — Goldziher · 2026-11-16
- Kind ICLR Reviewer Raises Score Despite Unresolved Concerns, Citing the Paper's Value — MarkSchmidtUBC · 2026-11-16
- Hand-editing 0.2% of weights plus 100 CPU-minutes turns an English TTS model into an 18MB Sanskrit one — kastnerkyle · 2026-11-16
- Why chirality matters for AI for Science: enantiomers behave differently in biology — madeofAjala · 2026-11-16
- RSS 2027 submissions open with FAQ clarifying two-stage review and negative-result bar — GeorgiaChal · 2026-11-16