Kaidi Xu

FaithCoT-Bench: Benchmarking Instance-Level Faithfulness of Chain-of-Thought Reasoning

UProp: Investigating the Uncertainty Propagation of LLMs in Multi-Step Decision-Making

EIP: Weighted Ranking of LLMs by Quantifying Question Difficulty

Beyond Redundancy: Diverse and Specialized Multi-Expert Sparse Autoencoder

Optimal Aggregation of LLM and PRM Signals for Efficient Test-Time Scaling

Symbiotic Cooperation for Web Agents: Harnessing Complementary Strengths of Large and Small LLMs

Calibrating the Voice of Doubt: How LLMs Diverge from Humans in Verbal Uncertainty

Spectral-Aware Sparse Communication and Entropy-Balanced Tasking in Multi-Agent Systems

Typography Leads Semantic Diversifying: Amplifying Adversarial Transferability across Multimodal Large Language Models

Optimizing Robustness and Accuracy in Mixture of Experts: A Dual-Model Approach

Adversarial Contrastive Decoding: Aligning Large Language Models via Exploiting Their Safety and Harm