Reinforcement learning with verifiable rewards (RLVR) has become an effective post-training paradigm for vision-language models (VLMs). Standard RLVR, however, assigns identical outcome-derived credit to all generated tokens regardless of their perceptual relevance. Recent perception-aware methods estimate visual relia...
A-RSI monitors reward variation and accuracy in training logs to refresh a task pool to refresh a task pool when its learning signal declines, without consulting validation or test scores.
Yu Liu, You-Xing Li, Yue-Dong Tan et al.· 0 citations
Multimodal large language models (MLLMs) answer questions about whole images well, yet they still struggle with fine-grained region understanding. A key reason is that text-only supervision never specifies which regions support an answer. To close this gap, we present Xiaomi-Lens-0, a 4B MLLM built on Qwen3.5 that pred...
OmniStepBench is introduced, a benchmark for deployment-aware agent evaluation across modalities and task complexity, and its findings support deployment decisions based on workload profiles, deployment mode, and accuracy--cost trade-offs rather than a single aggregate score.
Yu Liu, Zhi-Lin Liu, Zhi-Wei Yang et al.· 0 citations
An approximately 170M-sample OCR-centric corpus is built using an automated data engine that combines expert consensus, render-based verification, and targeted synthesis and shows that, with sufficient parsing training, OCR-centric understanding supervision provides additional gains for document parsing.
Xin Chen, An-An Du, Feng-Juan Feng et al.· 0 citations
The MaP is proposed, a unified framework that seamlessly harmonizes divergent GUI navigation tasks and effectively mitigates gradient conflicts and significantly outperforms the direct mixture training, establishing a robust paradigm for multi-task GUI navigation.
Yan Zhang, Pei Fu, Dai-Qing Wu et al.· 0 citations
DAREBench (Deployment-Aware and Reliable Evaluation of Models as Agents), a benchmark designed to capture workload variation and support reliable agent evaluation, is introduced, suggesting that agent deployment and model selection should consider workload profiles, deployment mode, and accuracy--cost trade-offs rather...
Yu Liu, Zhi-Lin Liu, Zhi-Wei Yang et al.· 0 citations
This work proposes NTEP-R (NTEP Reward), a supervision mechanism ensuring that each tool invocation strictly advances the reasoning process toward the final solution, and introduces a non-repeated-goal regularizer to penalize redundant calls that revisit satisfied NTEP goals.
Xing-Ming Long, Yu Liu, Zhi-Wei Yang et al.· 0 citations
Switch-Reasoner is proposed, a GRPO-based framework that learns to adaptively select reasoning modes for MLLMs and introduces a dual-level regulation mechanism that balances the overall use of Thinking Mode and Direct Mode while providing sample-level supervision based on the relative benefit of the two choices.
Yiyang Fang, Pei Fu, Jinjie Li et al.· arXiv.org· 0 citations
DeltaV is proposed, a ULMM that replaces full-image generation with visual updates and introduces a temporal similarity (TSIM) Router, which stops allocating tokens once the marginal reconstruction gain falls below a threshold.
Pengjie Wang, Linger Deng, Zujian Zhang et al.· 1 citation
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.