23 papers
Escaping Confidence Trap: Evolutionary Decoding for Mathematical Reasoning in Diffusion LLMs
Zhenhong Sun, Hanqing Zhao, Yatao Bian +7
Diffusion large language models (dLLMs) have emerged as a promising alternative to autoregressive LLMs, offering efficient generation through block-wise progressive unmasking. Howe…
Beyond One-Size-Fits-All: Diagnosis-Driven Online Reinforcement Learning with Offline Priors
Guozheng Ma, Lu Li, Zilin Wang +2
Online reinforcement learning (RL) agents increasingly depend on knowledge acquired offline to achieve practical efficiency. Originally studied in offline-to-online RL, this paradi…
Beyond Scalar Scores: Exploring LLM-based Metrics for Clinical Significance Evaluation in Radiology Reports
Qingyu Lu, Ruochen Li, Liang Ding +3
Reliable evaluation of generated radiology reports requires strict clinical accuracy, as omitted critical findings or mischaracterized radiographic observations can directly affect…
Attribution-Guided and Coverage-Maximized Pruning for Structural MoE Compression
Yifu Ding, Jiacheng Wang, Ge Yang +4
Mixture-of-Experts (MoE) models scale compute efficiently, yet remain expensive to deploy due to their substantial memory footprint and inference overhead. Prior compression method…
TimeGuard: Channel-wise Pool Training for Backdoor Defense in Time Series Forecasting
Quang Duc Nguyen, Siyuan Liang, Yiming Li +2
Time Series Forecasting (TSF) is highly vulnerable to backdoor attacks, yet effective defenses remain underexplored due to challenges arising from data entanglement and shifts in t…
ConceptSeg-R1: Segment Any Concept via Meta-Reinforcement Learning
Yuan Zhao, Youwei Pang, Jiaming Zuo +10
Recent progress in promptable segmentation has shifted visual perception from object-level localization toward concept-level understanding. However, the notion of a concept remains…
PixVerve: Advancing Native UHR Image Generation to 100MP with a Large-Scale High-Quality Dataset
Haojun Chen, Haoyang He, Chengming Xu +11
Text-to-Image (T2I) models have recently seen notable progress around 1K and 2K resolution. With the extreme desire for better visual experience and the rapid development of imagin…
SPIKE: An Adaptive Dual Controller Framework for Cost-Efficient Long-Horizon Game Agents
Wencan Jiang, Jiangning Zhang, Jianbiao Mei +6
Long-horizon multimodal agents in open-world games must stay goal-directed across many low-level interactions under tight token and latency budgets. Existing approaches often trade…
Who Generated This 3D Asset? Learning Source Attribution for Generative 3D Models
Sihan Ma, Siyuan Liang, Dacheng Tao
Generative 3D models are deployed in gaming, robotics, and immersive creation, making source attribution critical: given a 3D asset, can we identify whether and which generative mo…
STRIDE: Learnable Stepwise Language Feedback for LLM Reasoning
Junjie Zhang, Guozheng Ma, Shunyu Liu +5
Recent advances in Reinforcement Learning (RL) have underscored its potential for incentivizing reasoning capabilities of Large Language Models (LLMs). However, existing step-level…
Learn to Think: Improving Multimodal Reasoning through Vision-Aware Self-Improvement Training
Qihuang Zhong, Liang Ding, Wenjie Xuan +3
Post-training with explicit reasoning traces is common to improve the reasoning capabilities of Multimodal Large Language Models (MLLMs). However, acquiring high-quality reasoning…
SimReg: Achieving Higher Performance in the Pretraining via Embedding Similarity Regularization
Yan Sun, Guoxia Wang, Jinle Zeng +6
Pretraining large language models (LLMs) with next-token prediction has led to remarkable advances, yet the context-dependent nature of token embeddings in such models results in h…
WMAttack: Automated Attack Search for Adversarial Evaluation of World-Model Agents
Zhixiang Guo, Siyuan Liang, Shi Fu +4
Despite the growing use of world models as decision-making agents, their adversarial robustness remains underexplored due to the lack of dedicated automated evaluation methods. A k…
Distillation Traps and Guards: A Calibration Knob for LLM Distillability
Weixiao Zhan, Yongcheng Jing, Leszek Rutkowski +1
Knowledge distillation (KD) transfers capabilities from large language models (LLMs) to smaller students, yet it can fail unpredictably and also underpins model leakage risks. Our…
Multi-Domain Learning with Global Expert Mapping
Pourya Shamsolmoali, Masoumeh Zareapoor, Huiyu Zhou +3
Human perception generalizes well across different domains, but most vision models struggle beyond their training data. This gap motivates multi-dataset learning, where a single mo…
Understanding and Enforcing Weight Disentanglement in Task Arithmetic
Shangge Liu, Yuehan Yin, Lei Wang +5
Task arithmetic provides an efficient, training-free way to edit pre-trained models, yet lacks a fundamental theoretical explanation for its success. The existing concept of ``weig…
Try, Check and Retry: A Divide-and-Conquer Framework for Boosting Long-context Tool-Calling Performance of LLMs
Kunfeng Chen, Qihuang Zhong, Juhua Liu +2
Tool-calling empowers Large Language Models (LLMs) to interact with external environments. However, current methods often struggle to handle massive and noisy candidate tools in lo…
VideoTIR: Accurate Understanding for Long Videos with Efficient Tool-Integrated Reasoning
Zhe Gao, Shiyu Shen, Taifeng Chai +7
Existing Multimodal Large Language Models (MLLMs) often suffer from hallucinations in long video understanding (LVU), primarily due to the imbalance between textual and visual toke…
UniICL: Systematizing Unified Multimodal In-context Learning through a Capability-Oriented Taxonomy
Yicheng Xu, Jiangning Zhang, Zhucun Xue +5
In-context learning (ICL) enables fast task adaptation from demonstrations without per-task parameter updates but remains highly sensitive to example selection and formatting. In u…
EvoSyn: Generalizable Evolutionary Data Synthesis for Verifiable Learning
He Du, Bowen Li, Aijun Yang +3
Reliable verifiable data has become a key driver of capability gains in modern language models, enabling stable reinforcement learning with verifiable rewards and effective distill…
Confidence as a Reward: Transforming LLMs into Reward Models
He Du, Bowen Li, Chengxing Xie +3
Reward models can significantly enhance the reasoning capabilities of large language models (LLMs), but they typically require extensive curated data and costly training. To mitiga…
MaskPro: Linear-Space Probabilistic Learning for Strict (N:M)-Sparsity on LLMs
Yan Sun, Qixin Zhang, Zhiyuan Yu +3
The rapid scaling of large language models~(LLMs) has made inference efficiency a primary bottleneck in the practical deployment. To address this, semi-structured sparsity offers a…
GeometryZero: Advancing Geometry Solving via Group Contrastive Policy Optimization
Yikun Wang, Yibin Wang, Dianyi Wang +4
Recent progress in large language models (LLMs) has boosted mathematical reasoning, yet geometry remains challenging where auxiliary construction is often essential. Prior methods…