2 citations · 5 across the 19 of their papers we have counts for
21 papers
Register Tokens for Bounded-State Reasoning in Diffusion Language Models
Albert Ge, Chandan Singh, Yufan Zhuang +3
Masked diffusion language models (dLLMs) generate text by iteratively denoising masked tokens with bidirectional attention. Extending reasoning across generation chunks normally re…
Weak-to-Strong On-Policy Distillation
Fangxu Yu, Weijia Xu, Michael Xu +2
On-policy distillation (OPD), which aligns a student with the teacher's token-level distribution on the student's own rollouts, is an effective paradigm for transferring capabiliti…
GFlowRL: Scaling Distribution-Matching RL to Large Language Models
Xiaodong Liu, Michael Xu, Jack W. Stokes +3
Generative Flow Networks (GFlowNets) offer a promising alternative to reward-maximizing reinforcement learning (RL) for large reasoning models, encouraging diverse reasoning paths…
PolyAlign: Conditional Human-Distribution Alignment
L. D. M. S. Sai Teja, Ufaq Khan, Sathira Silva +2
Post-training methods such as supervised fine-tuning (SFT) and preference optimization typically align language models toward a single global assistant behavior. While effective fo…
MultiBreak: A Scalable and Diverse Multi-turn Jailbreak Benchmark for Evaluating LLM Safety
Jialin Song, Xiaodong Liu, Weiwei Yang +4
We present MultiBreak, a scalable and diverse multi-turn jailbreak benchmark to evaluate large language model (LLM) safety. Multi-turn jailbreaks mimic natural conversational setti…
Training Large Reasoning Models Efficiently via Progressive Thought Encoding
Zeliang Zhang, Xiaodong Liu, Hao Cheng +3
Large reasoning models (LRMs) excel on complex problems but face a critical barrier to efficiency: reinforcement learning (RL) training requires long rollouts for outcome-based rew…
Statistical Estimation of Adversarial Risk in Large Language Models under Best-of-N Sampling
Mingqian Feng, Xiaodong Liu, Weiwei Yang +3
Large Language Models (LLMs) are typically evaluated for safety under single-shot or low-budget adversarial prompting, which underestimates real-world risk. In practice, attackers…
SEMA: Simple yet Effective Learning for Multi-Turn Jailbreak Attacks
Mingqian Feng, Xiaodong Liu, Weiwei Yang +4
Multi-turn jailbreaks capture the real threat model for safety-aligned chatbots, where single-turn attacks are merely a special case. Yet existing approaches break under exploratio…
Multimodal Reinforcement Learning with Adaptive Verifier for AI Agents
Reuben Tan, Baolin Peng, Zhengyuan Yang +15
Agentic reasoning models trained with multimodal reinforcement learning (MMRL) have become increasingly capable, yet they are almost universally optimized using sparse, outcome-bas…
FlowRL: Matching Reward Distributions for LLM Reasoning
Xuekai Zhu, Daixuan Cheng, Dinghuai Zhang +20
We propose FlowRL: matching the full reward distribution via flow balancing instead of maximizing rewards in large language model (LLM) reinforcement learning (RL). Recent advanced…
SAS: Simulated Attention Score
Chuanyang Zheng, Jiankai Sun, Yihang Gao +12
The attention mechanism is a core component of the Transformer architecture. Various methods have been developed to compute attention scores, including multi-head attention (MHA),…
StreamAdapter: Efficient Test Time Adaptation from Contextual Streams
Dilxat Muhtar, Yelong Shen, Yaming Yang +11
In-context learning (ICL) allows large language models (LLMs) to adapt to new tasks directly from the given demonstrations without requiring gradient updates. While recent advances…
Generative Adapter: Contextualizing Language Models in Parameters with A Single Forward Pass
Tong Chen, Hao Fang, Patrick Xia +5
Large language models (LMs) are typically adapted to improve performance on new contexts (\eg text prompts that define new tasks or domains) through fine-tuning or prompting. Howev…
Iterative Self-Tuning LLMs for Enhanced Jailbreaking Capabilities
Chung-En Sun, Xiaodong Liu, Weiwei Yang +5
Recent research has shown that Large Language Models (LLMs) are vulnerable to automated jailbreak attacks, where adversarial suffixes crafted by algorithms appended to harmful quer…
GRIN: GRadient-INformed MoE
Liyuan Liu, Young Jin Kim, Shuohang Wang +14
Mixture-of-Experts (MoE) models scale more effectively than dense models due to sparse computation through expert routing, selectively activating only a small subset of expert modu…
Model Tells Itself Where to Attend: Faithfulness Meets Automatic Attention Steering
Qingru Zhang, Xiaodong Yu, Chandan Singh +6
Large language models (LLMs) have demonstrated remarkable performance across various real-world tasks. However, they often struggle to fully comprehend and effectively utilize thei…
Diversifying the Expert Knowledge for Task-Agnostic Pruning in Sparse Mixture-of-Experts
Zeliang Zhang, Xiaodong Liu, Hao Cheng +2
By increasing model parameters but activating them sparsely when performing a task, the use of Mixture-of-Experts (MoE) architecture significantly improves the performance of Large…
Towards Consistent Natural-Language Explanations via Explanation-Consistency Finetuning
Yanda Chen, Chandan Singh, Xiaodong Liu +4
Large language models (LLMs) often generate convincing, fluent explanations. However, different from humans, they often generate inconsistent explanations on different inputs. For…
Tell Your Model Where to Attend: Post-hoc Attention Steering for LLMs
Qingru Zhang, Chandan Singh, Liyuan Liu +4
In human-written articles, we often leverage the subtleties of text style, such as bold and italics, to guide the attention of readers. These textual emphases are vital for the rea…
Fast-ELECTRA for Efficient Pre-training
Chengyu Dong, Liyuan Liu, Hao Cheng +3
ELECTRA pre-trains language models by detecting tokens in a sequence that have been replaced by an auxiliary model. Although ELECTRA offers a significant boost in efficiency, its p…
ReEval: Automatic Hallucination Evaluation for Retrieval-Augmented Large Language Models via Transferable Adversarial Attacks
Xiaodong Yu, Hao Cheng, Xiaodong Liu +2
Despite remarkable advancements in mitigating hallucinations in large language models (LLMs) by retrieval augmentation, it remains challenging to measure the reliability of LLMs us…