Vision-language-action policies control embodied agents through a perception-action loop grounded in physical interaction, where adversarial patches can corrupt action predictions and thereby alter subsequent observations. To mitigate these threats, many existing defenses rely on policy adaptation or auxiliary models, but require additional training or inference. We introduce RECLAIM, a training-free, inference-time defense that uses a frozen VLA's own attention response to local visual interventions to guide purification during closed-loop embodied manipulation. Our key insight is that attention changes induced by local edits provide feedback for choosing where and how to purify as the robot interacts with its environment. RECLAIM combines this feedback with robot geometry to refine patch localization and select local treatments, then revalidates them as the agent's embodied actions change the observed scene. We evaluate RECLAIM against patches generated by four adversarial attack methods across multiple manipulation task suites under frozen VLA control. Under adversarial patch attacks, RECLAIM succeeds in at least one rollout on 95% of the evaluated tasks and raises rollout success on LIBERO-Object from 0% without defense to 70.0%. Its mean decision time is only 548.9 ms per control step. These results highlight policy-guided visual purification as an effective approach to defending embodied agents against adversarial patches and recovering closed-loop task execution.
Figures & tables
Figure 1: Policy-guided defense against adversarial patches. (a) A patch can redirect a frozen VLA’s actions and disrupt task execution. (b) Existing defenses update the VLA or invoke an auxiliary model. (c) RECLAIM uses the frozen policy’s own attention and action feedback to guide localization, purification, and temporal revalidation.
Figure 2: RECLAIM framework. GAL combines persistent attention with projected robot occupancy to propose regions. ARP evaluates fresh edits and recalled pairs on the current image, requiring cross-family support and ranking eligible treatments by attention reduction and bounded entropy gain. Protected selection yields a local edit or PASS for action generation by the frozen VLA. MTR stores selected pairs, recalls them at full decisions, and checks geometry and action validity between decisions.
Vision-Language-Action (VLA) policies promise general robotic manipulation, but their robustness against physical-world attacks remains fragile. In particular, we show that physically realizable adversarial patches can reliably induce failures by triggering a mechanism we call policy-critical action-to-vision attention hijacking, where action-conditioned attention is diverted from task-relevant regions to a localized patch. To demonstrate the threat, we propose Attention-Guided Semantic Disruption (AGSD), an Expectation-over-Transformation (EOT) optimized printable patch that jointly (i) concentrates action-to-vision attention on the patch and (ii) disrupts vision-language semantic alignment, yielding strong cross-task and cross-architecture transfer. To mitigate such attacks, we introduce Structure-Aware Robust Fine-Tuning (SARF), a zero-inference-overhead defense that fine-tunes only the visual encoder using feature anchoring, policy-critical attention correction, and language-guided geometric consistency restricted to semantically relevant regions. On LIBERO, SARF reduces OpenVLA's failure rate under AGSD from 100% to 14.2%-56.8% (28.6% average) across suites while preserving clean performance, and on a real PiPER manipulator it improves average success under AGSD from 23.0% to 65.0%. These results highlight mechanism-level robustness as a practical path to securing VLA robots against physical attention hijacking.
Jinquan Zhang, Dongfu Yin, Run Yang +3
Guangdong Laboratory of Artificial Intelligence and Digital Economy (SZ), Shenzhen, China · Shenzhen University, Shenzhen, China · Carleton University, Ottawa, Canada
Adversarial patches can disrupt Vision-Language-Action (VLA) models by manipulating visual observations, leading to failures in robot control. However, it remains poorly understood which internal mechanisms underlie these failures and how targeted interventions can mitigate them. In this work, we mechanistically analyze VLA representations using a sparse autoencoder (SAE) and identify a feature whose activation strongly correlates with the presence of an adversarial patch. Based on this analysis, we suppress the identified feature at inference time only when a linear probe detects an attack. This intervention improves robustness without the cost of fine-tuning the VLA. We evaluate our method against VLA adversarial patch attacks on LIBERO-10. Conditional intervention improves success rate under intermittent attacks, whereas continuously applying the same intervention substantially degrades policy performance. These results show that attack-related internal representations can provide useful targets for VLA adversarial defense and that controlling when to intervene is important for limiting disruption to nominal policy behavior.
Yukiya Horiba, Koshiro Aoki, Shunsuke Yasuki +2
Keio University · Waseda University · The University of Tokyo
Adversarial patches to Vision-Language-Action (VLA) policies can cause both immediate action corruption and persistent state effects that remain after the patch is removed. Existing evaluations largely focus on continuous attacks and do not separate these two effects. We introduce a state-restoration protocol that removes the patch at matched action-chunk boundaries and measures subsequent recoverability under the same remaining step budget. Clean, random-patch, deviation-matched, and fixed-direction controls distinguish adversarial effects from occlusion, action-error magnitude, and directional persistence. We also evaluate a recovery adapter trained on attack-induced states under controlled intervention latency. On OpenVLA-OFT with EDPA attacks, only 36.2% of LIBERO-Long episodes remain recoverable after five chunks, compared with 89.9% and 87.0% for the deviation-matched and fixed-direction controls. Similar persistent effects are observed on autoregressive OpenVLA. The recovery adapter improves recovery from 7.7% to 47.4% at one-chunk latency, but its benefit decreases substantially with delayed intervention. These results show that adversarial effects can persist after patch removal and that timely intervention is critical for recovery.
Enhao Wu, Fusen Guo, Yuxin Cao +3
UNSW Sydney, Australia · UNSW Canberra, Australia · National University of Singapore, Singapore +2