Built with and by Teycir Ben Soltane•
How to Use•FAQ•GitHub•arXiv.org•
Share:
ArXivCSExplorer
☆☆Bookmarks🏆RSSHow to UseFAQ
Home/Authors/Hanjie Chen

Hanjie Chen

3 indexed papers

Recent (6 mo)
3
With code
0
Influential cites
0
Benchmarked
0

Publications per year

3
26

Top categories

NLP×2AI×2ML×2Stats ML×1Crypto×1Multiagent×1

Frequent co-authors

Zilin Xiao1×
Qi Ma1×
Chun-cheng Jason Chen1×
Xintao Chen1×
Avinash Atreya1×
Vicente Ordonez1×

Research Timeline

2026
When Embedding-Based Defenses Fail: Rethinking Safety in LLM-Based Multi-Agent Systems

This paper analyzes the failure of current embedding-based defenses in multi-agent LLM systems and proposes using token-level confidence scores (logits) for improved robustness.

Conformal Certification of Reasoning Trace Prefixes

The paper introduces CROP, a novel conformal procedure that provides rigorous statistical guarantees for certifying the longest safe prefix of a language model's reasoning trace, allowing for targeted error identification and repair.

Learning to Reason by Analogy via Retrieval-Augmented Reinforcement Fine-Tuning

This paper proposes a post-training framework called Retrieval-Augmented Reinforcement Fine-Tuning (RA-RFT) to teach language models to reason by analogy.

Highlighted terms show continued research focus across papers

Papers

cs.CLcs.AIEmpiricalRecentJun 11, 2026

Learning to Reason by Analogy via Retrieval-Augmented Reinforcement Fine-Tuning

Zilin Xiao, Qi Ma, Chun-cheng Jason Chen, Xintao Chen +3 more

This paper proposes a post-training framework called Retrieval-Augmented Reinforcement Fine-Tuning (RA-RFT) to teach language models to reason by analogy.

View →
cs.AIcs.CLcs.LGRecent
May 28, 2026

Conformal Certification of Reasoning Trace Prefixes

Matt Y. Cheung, Ashok Veeraraghavan, Hanjie Chen, Guha Balakrishnan

The paper introduces CROP, a novel conformal procedure that provides rigorous statistical guarantees for certifying the longest safe prefix of a language model's reasoning trace, allowing for targeted…

View →
cs.CRcs.LGcs.MARecentMay 1, 2026

When Embedding-Based Defenses Fail: Rethinking Safety in LLM-Based Multi-Agent Systems

Lingxi Zhang, Guangtao Zheng, Hanjie Chen

This paper analyzes the failure of current embedding-based defenses in multi-agent LLM systems and proposes using token-level confidence scores (logits) for improved robustness.

View →