Vision-language-action (VLA) models transfer semantic knowledge acquired through vision-language pretraining to robot control, achieving strong performance across various benchmarks. However, most VLAs directly map observations to demonstrated actions without explicit supervision of intermediate physical state changes,...
Qian-Yu Lu, Jia-Run Li, Ce Zhang et al.· 0 citations
VLA-Rolling is introduced, a lightweight inference framework that reuses action tails through three stages: action-tail inheritance, per-action confidence estimation, and confidence-guided refinement through renoising and denoising budget allocation, while the underlying VLA remains frozen.
Si-Cheng He, Ying-Ying Pan, Qiu-Ting Yu et al.· 0 citations
Long-horizon embodied interaction requires agents to retain and continually update information about the environment as they observe, act, and encounter change. Yet current agents struggle to maintain such memory reliably. Our analysis traces this limitation to four key deficiencies: weak fine-grained visual memory, un...
Li-Zhou Liang, Xin-Yu Zhong, Miao Pan et al.· 0 citations
It is found that early OPD training dynamics uniformly exhibit a regular *useful-transfer* regime, in which held-out accuracy (the *gold score*, $G$) rises approximately linearly in $d=\sqrt{\mathrm{KL}(\pi_\theta \Vert \pi_{\mathrm{ref}})}$, the square root of token-level reverse KL divergence from the student initial...
Yun-Tai Bao, Qin-Feng Li, Guo-qing Jiang et al.· 0 citations
Dual-Entropy Enhanced Policy Optimization (DEEPO), a dual-stage enhancement combining signal variance regularization with gradient preconditioning with semantic-entropy-triggered expert prefixes inject grounded continuations on high-uncertainty queries, providing direct supervision and restoring advantage variance.
Ying-Xuan Zhuang, Miao Pan, Wang-Jie Gan et al.· 0 citations
Experiments with GRPO and GiGPO on ALFWorld, WebShop, and SearchQA show that both axes provide independent gains and that their combi- nation consistently achieves the strongest overall performance across model scales.
Ying-Xuan Zhuang, Bin-He Yu, Jing-Xiao Yang et al.· 1 citation
AttriMem is proposed, an attribution-guided process-feedback framework for learning memory-construction policies with RL that outperforms retrieval-based, heuristic, and RL-based baselines, generalizes across benchmarks and answer models, and stabilizes RL optimization.
TASPO, which converts privileged supervision into outcome-grounded action credit, is introduced and indicates that TASPO reduces supervision mismatch and that action-level assignment stabilizes the policy optimization process.
Jing-Xiao Yang, Wang-Jie Gan, Ying-Xuan Zhuang et al.· 4 citations
Experiments show that AttriMem outperforms retrieval-based, heuristic, and RL-based baselines, generalizes across benchmarks and answer models, stabilizes RL optimization, and outperforms retrieval-based, heuristic, and RL-based baselines on long-horizon dialogue question answering.
Qin-Feng Li, Yun-Tai Bao, Xinyang Yu et al.· 0 citations
SkillAligner is proposed, a training-free execution-time skill adaptation framework that treats retrieved skills as adaptable drafts rather than fixed instructions that substantially improves task performance over existing skill-use baselines, reduces skill-induced regressions at the instance level, and lowers total in...
Qin-Feng Li, Dalin He, Yun-Tai Bao et al.· 0 citations
This work proposes TAMP-Nav, a unified framework for efficient embodied navigation that dynamically triggers Chain-of-Thought and retains high-fidelity memory only at critical nodes, compressing redundant trajectories into lightweight Space-Time Indicators, thereby preserving critical historical information and enhanci...
Hongyan Feng, Sun-Lai Chen, Xuan-Yu Liu et al.· 1 citation
ProVisE (Protocolized Visual Evaluation), a benchmark-agnostic framework that elicits protocol-constrained visual answers from image-generation models and parses them into structured predictions compatible with original metrics, is proposed and revealed, revealing complementary strengths of pixel-space expression and t...
Xu Wang, Kaixiang Yao, Miao Pan et al.· arXiv.org· 1 citation
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
