285 citations · 736 across the 85 of their papers we have counts for
17 papers · 1 filter
When to Stop Reusing: Dynamic Gradient Gating for Sample-Efficient RLVR
Yuchun Miao, Sen Zhang, Yuqi Zhang +4
Reinforcement Learning with Verifiable Rewards (RLVR) has become the dominant paradigm for advanced reasoning in Large Language Models (LLMs), but rollout samples are expensive to…
Information-Theoretic Reward Modeling for Stable RLHF: Detecting and Mitigating Reward Hacking
Yuchun Miao, Liang Ding, Sen Zhang +3
Despite the success of Reinforcement Learning from Human Feedback (RLHF) in aligning language models with human values, reward hacking-or reward over-optimization-remains a major c…
OmniBridge: Unified Multimodal Understanding, Generation, and Retrieval via Latent Space Alignment
Teng Xiao, Zuchao Li, Lefei Zhang
Recent advances in multimodal large language models (LLMs) have led to significant progress in understanding, generation, and retrieval tasks. However, current solutions often trea…
Refining Few-Step Text-to-Multiview Diffusion via Reinforcement Learning
Ziyi Zhang, Li Shen, Deheng Ye +5
Text-to-multiview (T2MV) diffusion models have shown great promise in generating multiple views of a scene from a single text prompt. While few-step backbones enable real-time T2MV…
The Energy Loss Phenomenon in RLHF: A New Perspective on Mitigating Reward Hacking
Yuchun Miao, Sen Zhang, Liang Ding +3
This work identifies the Energy Loss Phenomenon in Reinforcement Learning from Human Feedback (RLHF) and its connection to reward hacking. Specifically, energy loss in the final la…
Efficient and Effective Weight-Ensembling Mixture of Experts for Multi-Task Model Merging
Li Shen, Anke Tang, Enneng Yang +6
Multi-task learning (MTL) leverages a shared model to accomplish multiple tasks and facilitate knowledge transfer. Recent research on task arithmetic-based MTL demonstrates that me…
SMILE: Zero-Shot Sparse Mixture of Low-Rank Experts Construction From Pre-Trained Foundation Models
Anke Tang, Li Shen, Yong Luo +5
Deep model training on extensive datasets is increasingly becoming cost-prohibitive, prompting the widespread adoption of deep model fusion techniques to leverage knowledge from pr…
FusionBench: A Unified Library and Comprehensive Benchmark for Deep Model Fusion
Anke Tang, Li Shen, Yong Luo +5
Deep model fusion is an emerging technique that unifies the predictions or parameters of several deep neural networks into a single better-performing model in a cost-effective and…
InfoRM: Mitigating Reward Hacking in RLHF via Information-Theoretic Reward Modeling
Yuchun Miao, Sen Zhang, Liang Ding +3
Despite the success of reinforcement learning from human feedback (RLHF) in aligning language models with human values, reward hacking, also termed reward overoptimization, remains…
Merging Multi-Task Models via Weight-Ensembling Mixture of Experts
Anke Tang, Li Shen, Yong Luo +3
Merging various task-specific Transformer-based models trained on different tasks into a single unified model can execute all the tasks concurrently. Previous methods, exemplified…
Sparse is Enough in Fine-tuning Pre-trained Large Language Models
Weixi Song, Zuchao Li, Lefei Zhang +2
With the prevalence of pre-training-fine-tuning paradigm, how to efficiently adapt the pre-trained model to the downstream tasks has been an intriguing issue. Parameter-Efficient F…
Neural Collapse Terminus: A Unified Solution for Class Incremental Learning and Its Variants
Yibo Yang, Haobo Yuan, Xiangtai Li +6
How to enable learnability for new classes while keeping the capability well on old classes has been a crucial challenge for class incremental learning. Beyond the normal case, lon…
Bidirectional Looking with A Novel Double Exponential Moving Average to Adaptive and Non-adaptive Momentum Optimizers
Yineng Chen, Zuchao Li, Lefei Zhang +2
Optimizer is an essential component for the success of deep learning, which guides the neural network to update the parameters according to the loss on the training set. SGD and Ad…
Robust and Discriminative Labeling for Multi-label Active Learning Based on Maximum Correntropy Criterion
Bo Du, Zengmao Wang, Lefei Zhang +2
Multi-label learning draws great interests in many real world applications. It is a highly costly task to assign many labels by the oracle for one instance. Meanwhile, it is also h…
Exploring Representativeness and Informativeness for Active Learning
Bo Du, Zengmao Wang, Lefei Zhang +4
How can we find a general way to choose the most suitable samples for training a classifier? Even with very limited prior information? Active learning, which can be regarded as an…
TLR: Transfer Latent Representation for Unsupervised Domain Adaptation
Pan Xiao, Bo Du, Jia Wu +3
Domain adaptation refers to the process of learning prediction models in a target domain by making use of data from a source domain. Many classic methods solve the domain adaptatio…
Multi-class Active Learning: A Hybrid Informative and Representative Criterion Inspired Approach
Xi Fang, Zengmao Wang, Xinyao Tang +1
Labeling each instance in a large dataset is extremely labor- and time- consuming . One way to alleviate this problem is active learning, which aims to which discover the most valu…