How Much Vision Does Multimodal Reasoning Need? Vision-Stripping for Multimodal Benchmarks
Under review
Vision-stripping for multimodal benchmarks and multimodal reasoning.
Publications
You can also find my full publication list on Google Scholar.
Research Articles
Every Act Has Its Price: Compressed Moral Composition in Frontier LLMs
Under review
A study of compressed moral composition in frontier LLMs.
CUADebug: Diagnosing and Repairing Computer-Use Agent Failures
Under review
A framework for diagnosing and repairing computer-use agent failures with a CUA-specific error taxonomy, benchmark, and tool-augmented debugger.
SeeingEye: Agentic Information Flow Unlocks Multimodal Reasoning in Text-only LLMs
arXiv preprint arXiv:2510.25092; under review
Agentic information flow for unlocking multimodal reasoning in text-only LLMs.
Where LLM Agents Fail and How They Can Learn From Failures
arXiv preprint arXiv:2509.25370
A study of LLM agent failures and how agents can learn from failed trajectories.
Conference Papers
Understanding and Mitigating the Bias Inheritance in LLM-based Data Augmentation on Downstream Tasks
Published in ACL 2026, Oral, 2026
A systematic study of how bias can be inherited through LLM-generated synthetic data and how mitigation strategies behave across tasks.
Artificial Leviathan: Exploring Social Evolution of LLM Agents Through the Lens of Hobbesian Social Contract Theory
Published in The First Workshop on AI Behavioral Science, ACM SIGKDD 2024, 2024
A simulated LLM agent society for studying emergent social contracts through Hobbesian social contract theory.