publications

2026

1. Evaluating Model Capabilities After Benchmark Saturation

   Morio Hamada, Jehyeok Yeon, Matteo Migliarini, and 2 more authors

   2026

   Preprint

2. 

   Securing Multimodal AI through Internal Information Decomposition

   Jehyeok Yeon, Hyeonjeong Ha, Qiusi Zhan, and 1 more author

   Proceedings of the 43rd International Conference on Machine Learning (ICML 2026) , 2026

   Spotlight

3. 

   InferenceBench: A Benchmark for Open-Ended LLM Inference Optimization by AI Agents

   Jehyeok Yeon, Ben Rank, and Maksym Andriushchenko

   ICML 2026 Agents in the Wild (AIWILD) Workshop , 2026

   Spotlight

   Website

4. 

   Certifying Robustness of Agent Tool-Selection Under Adversarial Attacks

   Jehyeok Yeon, Isha Chaudhary, and Gagandeep Singh

   ICLR 2026 Agentic AI in the Wild (AIWILD) Workshop , 2026

   Website

5. 

   GSAE: Graph-Regularized Sparse Autoencoders for Robust LLM Safety Steering

   Jehyeok Yeon, Federico Cinus, Yifan Wu, and 1 more author

   Proceedings of the 40th Conference on Neural Information Processing Systems (NeurIPS 2026) , 2026

6. 

   ResearchArena: Evaluating Sabotage and Monitoring in Automated AI R&D

   Ben Rank*, Lena Libon*, Jehyeok Yeon, and 5 more authors

   COLM 2026 Workshop on Agent Behavior * Equal contribution. , 2026

   Website

2025

1. 

   TRAP: Targeted Redirecting of Agentic Preferences

   Jehyeok Yeon*, Hangoo Kang*, and Gagandeep Singh

   Proceedings of the 39th Conference on Neural Information Processing Systems (NeurIPS 2025) , 2025

   arXiv
2. The Power of Friendship: Analyzing Leadership and Adversarial Attacks in Multi-Agent Collaboration

   Jehyeok Yeon, and Lawrence Angrave

   Poster accepted to ACM Collective Intelligence 2025; Non-archival , 2025