Current text-based person retrieval (TBPR) systems are typically built on English-pretrained vision-language backbones and evaluated on English-only, single-domain benchmarks. This setting limits their applicability to multilingual queries and encourages reliance on dataset-specific shortcuts, resulting in weak cross-d...
Hao-Tian Yang, Cheng Peng, Hao-Bo Wang et al.· 0 citations
StruPO is proposed, a structure-aware prompt editor that uses a calibrated scene graph as a shared representation for policy conditioning and image-grounded supervision, and improves the faithfulness--aesthetics trade-off, with consistent gains in relational fidelity.
Wen-Song Ye, Ya-Li Ye, Zhi-Qing Xiao et al.· 0 citations
A key challenge in on-policy distillation (OPD) is extracting effective supervision from teacher knowledge. Supervision from token probabilities may underutilize internal knowledge, while direct hidden-state matching forces the student to shift to the teacher's feature space and complicates cross-architecture transfer....
Kai Tang, Xing Zheng, Hao-Bo Wang et al.· 1 citation
Rubric-based reinforcement learning (RL) evaluates responses against explicit criteria to construct rewards, offering a viable way to train large language models on open-ended tasks. However, standard methods compress criterion-level feedback into scalar rewards and assign uniform token advantages, leaving structured r...
Guang-Cheng Zhu, Ying-Fan Ma, Xing Zheng et al.· 0 citations
On-policy self-distillation (OPSD) uses a self-teacher conditioned on privileged information to turn the sparse outcome signal of reinforcement learning with verifiable rewards (RLVR) into dense token-level supervision. Yet the teacher's advantage may depend on information unavailable to the student at inference, raisi...
Yu-Hang Yang, Ming-Xuan Xia, Chao-Nan Ye et al.· 0 citations
OrthoPurify is proposed, a more efficient method to purify backdoored model weights via one-step orthogonal projection, which reduces the attack success rate to near zero while preserving the original performance across diverse benchmarks, without retraining the backdoored model or introducing inference-time overhead.
Bo Yang, Hao Zhou, Zhi-Fang Zhang et al.· 0 citations
In Partial Label Learning (PLL), each instance is associated with a candidate label set, with exactly one label being true. While most studies implicitly assume balanced class distributions, real-world data often exhibit severe class imbalance distributions, leading to the Long-Tailed Partial Label Learning (LT-PLL) pr...
Xiangyu Ren, Mingxuan Xia, Guang-Cheng Zhu et al.· Proceedings of the Thirty-Fi...· 0 citations
This work derives a reformulated pixel-level contrastive learning objective by modeling feature distributions with the von Mises-Fisher distribution, and introduces a Reliability-aware Filter based on the vMF-derived metrics, which performs adaptive class-wise pixel reliability assessment to identify unreliable pixels...
Yi-Bo Wang, Rui-Kang Xu, Guang-Cheng Zhu et al.· Proceedings of the Thirty-Fi...· 0 citations
Criterion-Distilled Policy Optimization (CriPO), which enhances rubric-based RL via on-policy self-distillation via on-policy self-distillation, to simultaneously address both unexplored and suppressed criteria.
Mingxuan Xia, Yuhang Yang, Chaonan Ye et al.· arXiv.org· 3 citations· ⚡2
The results suggest that active, structure-preserving evidence seeking is a promising paradigm for open-domain table retrieval, and TableSeek, a structure-preserving agentic search framework for heterogeneous table corpora is presented.
Jia-Ming Tian, Li-Yao Li, Wen-Song Ye et al.· 0 citations
DE-Venus is presented, a unified framework for data-efficient RLVR that treats supervision as evolving state across data preparation and policy optimization and reduces annotation and training costs without sacrificing scalable RL execution.
Shen-Zhi Yang, Guang-Cheng Zhu, Kai Tang et al.· 1 citation
FCCA, which estimates the signed input--error cross-covariance, whitens it with diagonal Fisher moments, truncates it in the resulting local metric, maps the selected directions back, and applies thin QR to obtain stable core coordinates, shows that a carefully selected fixed span can recover most of the benefit of mov...
Wen-Song Ye, Zhan-Ming Shen, Zhiqing Xiao et al.· 1 citation· ⚡1
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.