3 citations · 10 across the 39 of their papers we have counts for
43 papers
Interpreting Model Reasoning through Behavioral Counterfactuals
Shawn Im, James Oldfield, Adel Bibi +1
Faithfully understanding the internal process that occurs as models form their chain of thought is an important problem with the rise in model capabilities. However, this remains c…
How Much Do Activation Monitors Benefit from Original Model Computations?
James Oldfield, Shawn Im, Jay J Culligan +3
One powerful way to safeguard AI systems is to monitor their internal computations--where signs of undesirable behavior might be more easily detectable than from models' inputs and…
LEASH: Uncertainty-Aware Weak-to-Strong Escalation and Oversight for Computer-Use Agents
Qingni Wang, Changdae Oh, Yuzhe Yang +4
Computer-use agents increasingly tackle long-horizon tasks, where running a frontier model at every step is expensive and a small open-weight model alone often fails. Weak-to-stron…
AIM: Agentic Idea Management for Automated Research
Hyeong Kyu Choi, Bhavana Dalvi Mishra, Jiefeng Chen +7
Frontier LLMs are increasingly used to automate scientific research through iterative search. We distinguish idea-driven search from solution-driven search and identify three core…
ExecCritic: Learn to Test, Test to Improve for Coding Agents
Leitian Tao, Baolin Peng, Haorui Wang +7
Execution feedback can guide coding agents toward correct repository repairs, but only when the tests capture the behavior requested by the issue. Agent-generated tests can encode…
Do Unified Multimodal Models Think in One Space? A Lens Through Cross-Branch Steering
Yu Wang, Sharon Li
Unified multimodal models (UMMs) aim to integrate understanding and generation within a single architecture, yet it remains unclear whether these capabilities share a unified and t…
TRACE: Turn-level Reward Assignment via Credit Estimation for Long-Horizon Agents
Leitian Tao, Baolin Peng, Wenlin Yao +5
Multi-turn agents solve complex tasks through extended sequences of tool interactions before producing a final answer, making credit assignment a fundamental challenge during post-…
Tracing Agentic Failure from the Flow of Success
Samuel Yeh, Yiwen Zhu, Shaleen Deep +1
Failure attribution for LLM-based agentic systems, i.e., identifying which steps in a failure trajectory caused the task to fail, is critical for debugging and improving these syst…
Do Multi-Agent LLMs Explore Each Other?
Hyeong Kyu Choi, Jiatong Li, Wendi Li +2
Exploration is essential for reliable autonomy in multi-agent systems, yet it remains unclear whether large language model (LLM) agents can explore effectively when interacting wit…
Multi-Head Recurrent Memory Agents
Jiatong Li, Samuel Yeh, Sharon Li
Recurrent memory agents extend LLMs to arbitrarily long contexts by iteratively consolidating input into a fixed-size memory window. Despite their scalability, these agents exhibit…
Neglected Free Lunch from Post-training: Progress Advantage for LLM Agents
Changdae Oh, Wendi Li, Seongheon Park +3
Process reward models enable fine-grained, step-level evaluation of LLMs, yet building them for agentic settings remains prohibitively difficult: long-horizon interactions, irrever…
OpenHalDet: A Unified Benchmark for Hallucination Detection across Diverse Generation Scenarios
Xinyi Li, Zhen Fang, Yongxin Deng +12
Hallucination detection is essential for the reliable deployment of large language models (LLMs). However, existing evaluations face two core challenges: inconsistent inference con…
Hide-and-Seek in Trajectories: Discovering Failure Signals for VLA Runtime Monitoring
Seongheon Park, Wendi Li, Changdae Oh +4
Vision-Language-Action (VLA) models enable robots to follow natural language instructions and generalize across diverse tasks, but they remain vulnerable to execution failures that…
Cyclical Entropy Eruption: Entropy Dynamics in Agent Reinforcement Learning
Wendi Li, Shawn Im, Sharon Li
Agentic large language models are increasingly used to solve real-world tasks by reasoning over goals, invoking tools, and interacting with external environments. Reinforcement lea…
DECOR: Auditing LLM Deception via Information Manipulation Theory
Linyue Cai, Samuel Yeh, Jwala Dhamala +2
Large language models can deceive by subtly manipulating truthful information -- omitting key facts, shifting focus, or obscuring meaning -- making such behavior difficult to detec…
TorchUMM: A Unified Multimodal Model Codebase for Evaluation, Analysis, and Post-training
Yinyi Luo, Wenwen Wang, Hayes Bai +6
Recent advances in unified multimodal models (UMMs) have led to a proliferation of architectures capable of understanding, generating, and editing across visual and textual modalit…
Auditing Agent Harness Safety
Chengzhi Liu, Yichen Guo, Yepeng Liu +8
LLM agents increasingly run inside execution harnesses that dispatch tools, allocate resources, and route messages between specialized components. However, a harness can return a c…
On Safety Risks in Experience-Driven Self-Evolving Agents
Weixiang Zhao, Yichen Zhang, Yingshuo Wang +8
Experience-driven self-evolution has emerged as a promising paradigm for improving the autonomy of large language model agents, yet its reliance on self-curated experience introduc…
Why Multimodal In-Context Learning Lags Behind? Unveiling the Inner Mechanisms and Bottlenecks
Yu Wang, Sharon Li
In-context learning (ICL) enables models to adapt to new tasks via inference-time demonstrations. Despite its success in large language models, the extension of ICL to multimodal s…
Can LLMs Learn to Reason Robustly under Noisy Supervision?
Shenzhi Yang, Guangcheng Zhu, Bowen Song +7
Reinforcement Learning with Verifiable Rewards (RLVR) effectively trains reasoning models that rely on abundant perfect labels, but its vulnerability to unavoidable noisy labels du…
Combating Data Laundering in LLM Training
Muxing Li, Zesheng Ye, Sharon Li +1
Post-hoc unauthorized-training data detection for large language models (LLMs) typically assumes a query-with-originals regime: rights holders query a target LLM with raw proprieta…
LAD: Learning Advantage Distribution for Reasoning
Wendi Li, Sharon Li
Current reinforcement learning objectives for large-model reasoning primarily focus on maximizing expected rewards. This paradigm can lead to overfitting to dominant reward signals…
VAUQ: Vision-Aware Uncertainty Quantification for LVLM Self-Evaluation
Seongheon Park, Changdae Oh, Hyeong Kyu Choi +2
Large Vision-Language Models (LVLMs) frequently hallucinate, limiting their safe deployment in real-world applications. Existing LLM self-evaluation methods rely on a model's abili…
How Retrieved Context Shapes Internal Representations in RAG
Samuel Yeh, Sharon Li
Retrieval-augmented generation (RAG) enhances large language models (LLMs) by conditioning generation on retrieved external documents, but the effect of retrieved context is often…