Jun Yan
4 papers · Latest:
Natural Language Processing
RubricEM: Meta-RL with Rubric-guided Policy Decomposition beyond Verifiable Rewards
RubricEM is a meta-RL framework that uses rubrics to guide policy decomposition and reflection for training research agents without verifiable rewards.
2605.10899
Artificial IntelligenceSkillOS: Learning Skill Curation for Self-Evolving Agents
SkillOS enables LLM agents to self-evolve by learning to curate reusable skills from experience, outperforming baselines in various tasks.
2605.06614
Natural Language ProcessingGenerating Effective CoT Traces for Mitigating Causal Hallucination
This paper introduces a pipeline to generate Chain-of-Thought traces and a new metric (CHR) to mitigate causal hallucination in smaller LLMs.
2604.12748
Populations & EvolutionSynonymous Codon Usage Bias Overrides Phylogeny to Reflect Convergent Frond Architecture in a Rapidly Radiating Fern Family Thelypteridaceae
Ferns show that synonymous codon usage bias (CUB) can override phylogeny, reflecting convergent frond architecture driven by specific photosynthesis genes.
2604.03028
📬 Weekly AI Paper Digest
Get the top 10 AI/ML arXiv papers from the week — summarized, scored, and delivered to your inbox every Monday.