Skip to content
Preprint

UniMPA: A Unified Memory-Prediction-Action Model via Action-Grounded Transition Modeling

Sep 2026 · 3 citations · 113 references
Computer Science

TL;DR

UniMPA, a Unified Memory-Prediction-Action model that addresses transition ambiguity by modeling the intended future state evolution through a shared action-grounded transition interface, is proposed.

Abstract

Recent advances in Vision-Language-Action (VLA) models have improved robotic manipulation, yet observation-to-action learning remains limited by a fundamental transition realizability gap, manifested in three tightly coupled problems: (i) Transition ambiguity. Visually similar current observations may correspond to different manipulation phases and imply different subsequent transitions. (ii) Prediction--execution mismatch. A visually plausible predicted future observation does not necessarily correspond to a physically realizable transition. (iii) Experience--realization mismatch. A historically executable action pattern may not necessarily realize the intended transition in the current scene and therefore requires context-aware adaptation. Accordingly, we propose UniMPA, a Unified Memory-Prediction-Action model that addresses these problems through a shared action-grounded transition interface. (i) UniMPA introduces Persistent-Selective Future Prediction to resolve transition ambiguity by modeling the intended future state evolution. A persistent latent stream continuously tracks task-level progress, while a transition-critical pixel stream selectively resolves fine-grained interaction changes through memory-grounded prediction. (ii) To assess the physical executability of the anticipated transition, the predicted transition queries a temporal Visual-Action Memory Bank. The bank retrieves historically realized visual-action experience, grounding future prediction in executable evidence. (iii) To adapt executable experience to the current scene, an Action-Visual Memory Bank retrieves visually grounded action prototypes from historical action evolution. Prototype-Biased Flow then shifts the flow source toward a historically supported action manifold for context-aware refinement.

View source

Similar papers

#machine learning Preprint Aug 2026

LM-X: Explainable Vision--Language--Action Modeling via Progress, Event, and Uncertainty Prediction

LM-X is introduced, which organizes prediction across task, event, and motor scales without claiming anatomical correspondence and shows that explicit multi-timescale predictive state can strengthen control while exposing interpretable internal estimates.

Jin Lou, Zhi Jing, Xu-Peng Wang et al. · 0 citations
Preprint Aug 2026

HarnessWAM: Bridging Prediction and Deliberation in World Action Models

Results demonstrate that model-external structured state maintenance and closed-loop agentic decision making can effectively extend the local control capabilities of WAMs into embodied task execution that is plannable, verifiable, and recoverable.

Zhao-Peng Gu, Bing-Ke Zhu, Tianxin Lin et al. · 5 citations
Preprint Oct 2026

World-Calibrated Proposal-to-Action Flow for Vision-Language-Action Models

ProAct is introduced, a world-calibrated proposal-to-action framework that makes the generative source itself predictable and improves performance across simulation and real-world tasks while reducing denoising steps by 50%, inference latency by up to 25.8%, and increasing throughput by up to 34.8%.

Jie He, Wei Li, Jun-Wen Tong et al. · 0 citations
Preprint Aug 2026

SLIM-0.5B: Learning Action-Grounded Predictive Latents for Robot Manipulation

SLIM (Self-supervised Latent Interaction Model), a compact 0.5B-parameter latent interaction policy, which matches or exceeds representative large-scale VLA and world-action-model baselines with fewer parameters, no additional embodied pretraining, lower inference latency, and substantially lower GPU memory usage.

Jing-Kai Wang, Zihan Tang, Gu Zhang et al. · 0 citations
Sep 2026

SmoLSTM: A Compact Vision-Language-Action Model with Recurrent Memory that Persists

SmoLSTM couples a frozen 256M-parameter SmolVLM backbone to a matrix-memory LSTM control layer in which observation tokens and action queries are unified in a single causal stream that is never reset throughout the entire episode, capturing the episode in a recurrent state.

Jan-Gerrit Habekost, Parsa Mastouri Kashani, Connor Gäde et al. · 0 citations
Preprint Sep 2026

Magic-W0: A Structured World-Action Foundation Model for Physical Intelligence

World-action models (WAMs) augment robot policies with action-conditioned environment dynamics, yet existing approaches largely rely on future observation reconstruction or generic latent prediction and lack structured, control-oriented world representations tightly coupled with action generation. We introduce Magic-W0...

Xu-Hua Chen, Zhen-Han Yin, Yu Zhang et al. · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.