43 citations · 108 across the 70 of their papers we have counts for
10 papers · 1 filter
Attention Sinks and Outliers in Attention Residuals
Haozheng Luo, Haoran Dai, Shaoyang Zhang +10
We propose OASIS, an outlier- and sink-aware method that stabilizes dual-normalized attention-residual architectures through explicit null routing and token-to-depth null coupling.…
Balancing the Reasoning Load: Difficulty-Differentiated Policy Optimization with Length Redistribution for Efficient and Robust Reinforcement Learning
Yinan Xia, Haotian Zhang, Huiming Wang
Large Reasoning Models (LRMs) have shown exceptional reasoning capabilities, but they also suffer from the issue of overthinking, often generating excessively long and redundant an…
When One Modality Rules Them All: Backdoor Modality Collapse in Multimodal Diffusion Models
Qitong Wang, Haoran Dai, Haotian Zhang +2
While diffusion models have revolutionized visual content generation, their rapid adoption has underscored the critical need to investigate vulnerabilities, e.g., to backdoor attac…
Believe Your Model: Distribution-Guided Confidence Calibration
Xizhong Yang, Haotian Zhang, Huiming Wang +1
Large Reasoning Models have demonstrated remarkable performance with the advancement of test-time scaling techniques, which enhances prediction accuracy by generating multiple cand…
Efficient RLVR Training via Weighted Mutual Information Data Selection
Xinyu Zhou, Boyu Zhu, Haotian Zhang +2
Reinforcement learning (RL) plays a central role in improving the reasoning and alignment of large language models, yet its efficiency critically depends on how training data are s…
CodeScaler: Scaling Code LLM Training and Test-Time Inference via Reward Models
Xiao Zhu, Xinyu Zhou, Boyu Zhu +5
Reinforcement Learning from Verifiable Rewards (RLVR) has driven recent progress in code large language models by leveraging execution-based feedback from unit tests, but its scala…
Tree Training: Accelerating Agentic LLMs Training via Shared Prefix Reuse
Jinghui Wang, Shaojie Wang, Yinghan Cui +9
Agentic large language model (LLM) training often involves multi-turn interaction trajectories that branch into multiple execution paths due to concurrent tool use, think-mode, sub…
SeamlessFlow: A Trainer Agent Isolation RL Framework Achieving Bubble-Free Pipelines via Tag Scheduling
Jinghui Wang, Shaojie Wang, Yinghan Cui +24
We introduce SeamlessFlow, a server based reinforcement learning (RL) framework that addresses two core challenges in industrial scale RL: (1) decoupling RL training from the compl…
SRPO: A Cross-Domain Implementation of Large-Scale Reinforcement Learning on LLM
Xiaojiang Zhang, Jinghui Wang, Zifei Cheng +14
Recent advances of reasoning models, exemplified by OpenAI's o1 and DeepSeek's R1, highlight the significant potential of Reinforcement Learning (RL) to enhance the reasoning capab…
CLoG: Benchmarking Continual Learning of Image Generation Models
Haotian Zhang, Junting Zhou, Haowei Lin +6
Continual Learning (CL) poses a significant challenge in Artificial Intelligence, aiming to mirror the human ability to incrementally acquire knowledge and skills. While extensive…