14 papers
Combating Catastrophic Forgetting in Continual Domain Adaptation via Knowledge Recasting
Zijie Liu, Zhen Tan, Charles Fleming +1
Adapting language models to new domains during post-training risks degrading previously acquired capabilities, a phenomenon commonly known as catastrophic forgetting. Existing miti…
Sign-Constrained Intervention Effects for Domain-Generalizable ICU World Models
Zhen Xu, Nicholas Konz, Zhen Tan +2
Predicting how a patient's vital signs respond to an intervention is a central question in intensive care. World models can do so by learning dynamics as a function of prior action…
Do More Modalities Always Help? A Geometric Perspective on Missing-Modality Robustness
Songyuan Sui, Zhen Tan, Mohan Zhang +3
Missing modality remains a longstanding challenge in multimodal learning. Existing methods typically address this issue through modality recovery or adaptive strategies. However, t…
OPD-Aha: From Linguistic Momentum to Visual Reflection in Multimodal On-Policy Distillation
Chenhao Qiu, Dawei Li, Yechao Zhang +2
Privileged on-policy distillation improves multimodal reasoning by allowing a teacher to evaluate student trajectories using rich, training-only visual evidence. Both models score…
Does AI Reviewer See the Full Picture? Attacking and Defending Multimodal Peer Review
Xinyu Zhao, Rana Muhammad Shahroz Khan, Zhen Xu +2
The integration of Large Language Models (LLMs) and Multimodal LLMs (MLLMs) into scientific peer-review workflows introduces novel and significant risks for adversarial manipulatio…
FairGen: Preference-Aligned Diffusion for Demographically Equitable Medical Image Synthesis
Zhimin Li, Ruichen Zhang, Zhen Tan +3
Medical imaging is central to modern diagnostics, and artificial intelligence (AI) systems are increasingly used to support image-based analysis by improving efficiency, accuracy,…
Measuring Real-World Prompt Injection Attacks in LLM-based Resume Screening
Mohan Zhang, Yuqi Jia, Zhen Tan +4
LLMs are vulnerable to prompt injection attacks. However, this vulnerability has been primarily demonstrated conceptually in academic studies or through a few anecdotal case studie…
TRUST: A Framework for Decentralized AI Service v.0.1
Yu-Chao Huang, Zhen Tan, Mohan Zhang +3
Large Reasoning Models (LRMs) and Multi-Agent Systems (MAS) in high-stakes domains demand reliable verification, yet centralized approaches suffer four limitations: (1) Robustness,…
TMS: Trajectory-Mixed Supervision for Reward-Free, On-Policy SFT
Rana Muhammad Shahroz Khan, Zijie Liu, Zhen Tan +2
Reinforcement Learning (RL) and Supervised Fine-Tuning (SFT) are the two dominant paradigms for enhancing Large Language Model (LLM) performance on downstream tasks. While RL gener…
Model Editing as a Double-Edged Sword: Steering Agent Ethical Behavior Toward Beneficence or Harm
Baixiang Huang, Zhen Tan, Haoran Wang +6
Agents based on Large Language Models (LLMs) have demonstrated strong capabilities across a wide range of tasks. However, deploying LLM-based agents in high-stakes domains comes wi…
DOGe: Defensive Output Generation for LLM Protection Against Knowledge Distillation
Pingzhi Li, Zhen Tan, Mohan Zhang +3
Large Language Models (LLMs) represent substantial intellectual and economic investments, yet their effectiveness can inadvertently facilitate model imitation via knowledge distill…
The Quest for Efficient Reasoning: A Data-Centric Benchmark to CoT Distillation
Ruichen Zhang, Rana Muhammad Shahroz Khan, Zhen Tan +3
Data-centric distillation, including data augmentation, selection, and mixing, offers a promising path to creating smaller, more efficient student Large Language Models (LLMs) that…
Efficient MAP Estimation of LLM Judgment Performance with Prior Transfer
Huaizhi Qu, Inyoung Choi, Zhen Tan +6
LLM ensembles are widely used for LLM judges. However, how to estimate their accuracy, especially in an efficient way, is unknown. In this paper, we present a principled maximum a…
: Breaking Pragmatic Multi-Agent LLM Systems with Optimized Prompt Attacks
Rana Muhammad Shahroz Khan, Zhen Tan, Sukwon Yun +2
Most discussions about Large Language Model (LLM) safety have focused on single-agent settings but multi-agent LLM systems now create novel adversarial risks because their behavior…