EventVLA: Event-Driven Visual Evidence Memory for Long-Horizon Vision-Language-Action Policies

Hugging Face Daily Papers Papers
vision-language-action long-horizon-manipulation memory-augmented keyframe-evidence robotic-manipulation visual-evidence-memory

Summary

EventVLA introduces a sparse visual evidence memory framework for long-horizon robotic manipulation, achieving an average success rate improvement of +40% over state-of-the-art memory-augmented VLAs.

Memory remains a critical bottleneck for long-horizon robotic manipulation, as standard Vision-Language-Action (VLA) policies often fail when task-relevant cues become occluded or unobservable over time. While existing memory-augmented methods utilize historical context, they either suffer from severe information bottlenecks, incur high latency via decoupled dual systems, or rely on unselective buffers that accumulate massive visual redundancies. To address these limitations, we introduce EventVLA, an end-to-end framework founded on the concept of sparse visual evidence memory that comprises two core components: foundational visual anchors to retain initial and short-term contexts, and a dynamic Keyframe Evidence Memory (KEM) module. Specifically, KEM directly predicts future keyframe probabilities from the VLA's latent embeddings to autonomously capture and store sparse, task-critical visual events. This foresight-driven mechanism empowers the policy to dynamically evaluate the future causal utility of current observations, preserving transient visual evidence before it becomes unobservable. Furthermore, we propose RoboTwin-MeM, a diagnostic benchmark specifically designed to evaluate non-Markovian manipulation tasks with interactive visual evidence. Extensive evaluations show that across 17 memory-requiring simulation tasks and 4 real-world bimanual tasks, EventVLA achieves an average success rate improvement of +40% over state-of-the-art memory-augmented VLAs.
Original Article
View Cached Full Text

Cached at: 06/24/26, 09:47 AM

Paper page - EventVLA: Event-Driven Visual Evidence Memory for Long-Horizon Vision-Language-Action Policies

Source: https://huggingface.co/papers/2606.20092 Authors:

,

,

,

,

,

,

,

,

,

,

,

Abstract

EventVLA addresses long-horizon robotic manipulation challenges by introducing a sparse visual evidence memory framework with visual anchors and dynamic Keyframe Evidence Memory module for improved task performance.

Memory remains a critical bottleneck for long-horizon robotic manipulation, as standardVision-Language-Action(VLA) policies often fail when task-relevant cues become occluded or unobservable over time. While existingmemory-augmented methodsutilize historical context, they either suffer from severe information bottlenecks, incur high latency via decoupled dual systems, or rely on unselective buffers that accumulate massive visual redundancies. To address these limitations, we introduce EventVLA, an end-to-end framework founded on the concept of sparsevisual evidencememory that comprises two core components: foundationalvisual anchorsto retain initial and short-term contexts, and a dynamicKeyframe Evidence Memory(KEM) module. Specifically, KEM directly predicts future keyframe probabilities from the VLA’slatent embeddingsto autonomously capture and store sparse, task-critical visual events. This foresight-driven mechanism empowers the policy to dynamically evaluate the futurecausal utilityof current observations, preserving transientvisual evidencebefore it becomes unobservable. Furthermore, we propose RoboTwin-MeM, adiagnostic benchmarkspecifically designed to evaluatenon-Markovian manipulation taskswith interactivevisual evidence. Extensive evaluations show that across 17 memory-requiring simulation tasks and 4 real-world bimanual tasks, EventVLA achieves an average success rate improvement of +40% over state-of-the-art memory-augmented VLAs.

View arXiv pageView PDFProject pageGitHub14Add to collection

Get this paper in your agent:

hf papers read 2606\.20092

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2606.20092 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2606.20092 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2606.20092 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

IntentVLA: Short-Horizon Intent Modeling for Aliased Robot Manipulation

Hugging Face Daily Papers

IntentVLA is a history-conditioned visual-language-action framework that improves robot imitation learning stability by encoding short-horizon intents from visual observations, addressing challenges from partial observability and ambiguous observations. It also introduces AliasBench, an ambiguity-aware benchmark for evaluating such methods.

LabVLA: Grounding Vision-Language-Action Models in Scientific Laboratories

Hugging Face Daily Papers

LabVLA is a vision-language-action model for scientific laboratory automation, trained with a two-stage approach combining action token pretraining and flow matching. It achieves state-of-the-art success rates on the LabUtopia benchmark by leveraging simulated data to bridge the gap between household demonstrations and lab-specific tasks.

AtlasVA: Self-Evolving Visual Skill Memory for Teacher-Free VLM Agents

Hugging Face Daily Papers

AtlasVA is a teacher-free visual skill memory framework for vision-language model agents that uses spatial heatmaps, visual exemplars, and symbolic text skills to improve spatial decision-making in long-horizon tasks, outperforming baselines on several benchmarks.