AI Paper Atlas

Research map / Vision-language and multimodal / Robot action and navigation

Phase-Adaptive Fusion: Spatio-Temporal Modulation for VLA Models

Lueke Ni, Junhao Zhu, Xinyu Zhang

NeurIPS 2026 · Vision-language and multimodal › Robot action and navigation · approach: Vision-language-action foundation models

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes Phase-Adaptive Fusion, a spatio-temporal modulation method for VLA models that counters a state-dominant shortcut caused by modality imbalance when adapting to dynamic robotic environments.

Problem

VLA policies adapting to dynamic physical environments tend to over-rely on low-dimensional proprioceptive states (a state-dominant shortcut) because they offer a more direct path to loss minimization, causing modality imbalance against visual inputs.

Approach

Modulates fusion of visual and proprioceptive modalities across space and time (phase-adaptive) to prevent the policy from taking the state-dominant shortcut; the abstract does not specify the mechanism in detail.

Surprising findings

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy