Yude Zou, Yifei Yao, Guanbo Jiao, Jianming Ma, Guanjie Zheng, Xing Gao · 2026 · arXiv
DOI: 10.48550/arXiv.2610.26813 · open-access PDF
World Action Models (WAM) bersama-sama memprediksi video masa depan dan menghasilkan tindakan, namun tetap sensitif terhadap gangguan visual yang tidak relevan dengan tugas, yaitu, perubahan dalam penampilan visual yang tidak harus mempengaruhi perilaku robot. Kami mengidentifikasi kesenjangan dalam pelatihan WAM konvensional: prediksi video mendorong representasi untuk mempertahankan informasi penampilan, sementara generasi aksi tidak secara eksplisit memaksakan invariance terhadap penampilan yang tidak relevan dengan tugas.
Machine translation (Meta M2M100); the English original follows.
World Action Models (WAMs) jointly predict future video and generate actions, yet remain sensitive to task-irrelevant visual perturbations, i.e., changes in visual appearance that should not affect robot behavior. We identify a gap in conventional WAM training: video prediction encourages representations to preserve appearance information, while action generation does not explicitly enforce invariance to task-irrelevant appearance. Data augmentation exposes WAMs to such perturbations but does not require perturbed and nominal observations to produce consistent action representations, and align
Read in: English · Spanish · Chinese · Hindi · Arabic · Portuguese · French · German · Japanese · Russian · Korean · Indonesian · Vietnamese
GET https://store.agentexchange.work/gw/papers/verify?title=...&year=... (0.003 USD), fetch this record as JSON with /gw/papers/paper?id=2610.26813, or list what it cites with /gw/papers/references?id=.... Pay per call in USDC (x402) or with a 19 USD card key for 1,000 calls. Catalogue · llms.txtSources: OpenAlex (CC0), arXiv, Crossref. Agent Exchange