12 papers
TVIR: Building Deep Research Agents Towards Text-Visual Interleaved Report Generation
Xinkai Ma, Zhiqi Bai, Dingling Zhang +21
Deep Research Agents have shown strong capability in multi-step information retrieval, reasoning, and long-form report generation, but existing benchmarks and systems remain predom…
TRL-Bench: Standardizing Cross-Paradigm Representation-Level Evaluation of Tabular Encoders
Wei Pang, Xiangru Jian, Hehan Li +10
Tabular encoders are usually evaluated inside task-specific end-to-end pipelines, so models from different training paradigms are difficult to compare directly even when they opera…
SG-OPD: Sign-Gated On-Policy Distillation via Sign-Consistency Gating and Phased Teacher Sampling
Haoran Xu, Hongyu Wang, Yifei Gao +3
On-policy distillation (OPD) trains a student on its own trajectories with dense per-token supervision from a stronger teacher, and often outperforms off-policy distillation and st…
Visual Para-Thinker++: A Single-Policy Multi-Agent Framework for Visual Reasoning
Haoran Xu, Hongyu Wang, Yifei Gao +4
Visual reasoning requires integrating evidence distributed across regions, attributes, and relations, making single-chain reasoning prone to early perceptual commitment and halluci…
Trajectory-Refined Distillation
Li Jiang, Haoran Xu, Yichuan Ding +1
On-policy distillation (OPD) has become a central post-training tool for large language models (LLMs), providing dense per-token teacher supervision along the student's own rollout…
TD-Grokking: Learning from Zero-Reward Problems by Training-Time Decomposition
Ningyuan Xi, Hao Xu, Hongsheng Xin +1
Large language models (LLMs) have made remarkable progress in reasoning tasks, largely driven by post-training paradigms, especially reinforcement learning with verifiable rewards…
Toward Robust In-Context Learning: Leveraging Out-of-distribution Proxies for Target Inaccessible Demonstration Retrieval
Hao Xu, Rite Bo, Fausto Giunchiglia +2
Although studies have demonstrated that Large Language Models (LLMs) can perform well on Out-of-Distribution (OOD) tasks, their advantage tends to diminish as the distribution shif…
Federated Balanced Learning
Jiaze Li, Haoran Xu, Wanyi Wu +9
Federated learning is a paradigm of joint learning in which clients collaborate by sharing model parameters instead of data. However, in the non-iid setting, the global model exper…
Visual Para-Thinker: Divide-and-Conquer Reasoning for Visual Comprehension
Haoran Xu, Hongyu Wang, Jiaze Li +5
Existing LLM test-time scaling laws emphasize the emergence of self-reflective behaviors through extended reasoning length. Nevertheless, this vertical scaling strategy often encou…
Video-OPD: Efficient Post-Training of Multimodal Large Language Models for Temporal Video Grounding via On-Policy Distillation
Jiaze Li, Hao Yin, Haoran Xu +6
Reinforcement learning has emerged as a principled post-training paradigm for Temporal Video Grounding (TVG) due to its on-policy optimization, yet existing GRPO-based methods rema…
Federated Joint Learning for Domain and Class Generalization
Haoran Xu, Jiaze Li, Jianzhong Ju +1
Efficient fine-tuning of visual-language models like CLIP has become crucial due to their large-scale parameter size and extensive pretraining requirements. Existing methods typica…
Vision Also You Need: Navigating Out-of-Distribution Detection with Multimodal Large Language Model
Haoran Xu, Yanlin Liu, Zizhao Tong +8
Out-of-Distribution (OOD) detection is a critical task that has garnered significant attention. The emergence of CLIP has spurred extensive research into zero-shot OOD detection, o…