Causal Analysis of Robot Foundation Models Enables Efficiency Gains

Aiden Swann, Mathilde Papillon, Mozes Jacobs, Siddharth Boppana, Tal Haklay, Eric Bigelow, Andrew Lee, Conor Watts, Thomas Icard, Mac Schwager, Owen Lewis, Jack Merullo, Ekdeep Singh Lubana, Atticus Geiger, Thomas Fel, Matthew Kowal · N/A · 2026

Summary

Vision-language-action (VLA) models are a promising approach to general purpose robot manipulation.

Abstract Summary

Vision-language-action (VLA) models are a promising approach to general purpose robot manipulation. While the internal computation of vision-language models, which compose the backbone of VLAs, has been studied extensively, less is known about how their perceptual and semantic representations are transformed into action in VLA models. We study this question mechanistically in three frontier VLA models ($\pi_{0.5}$, MolmoAct2, and MolmoBot) by analyzing the causal pathways through which the input image and the instruction shape actions. Across the three models, activation patching shows that causally relevant information influences the action expert (AE) through a narrow band of mid-to-late cross-attention layers, which we term the action-conditioning interface. The content of this information changes with token and layer within the model. Early-to-mid layer representations of the target noun token determine which object is selected, whereas mid-to-late layers carry the spatial information in text tokens that redirects the movement when patched. Where spatial information is stored in prompt tokens various based on the architecture. In MolmoAct2 the model forms task-conditioned representations at the text positions that follow the instruction, whereas $\pi_{0.5}$ additionally stores this information at image token positions. We show that these findings can be translated directly into improving model efficiency. We find that restricting MolmoBot's action expert conditioning to only 3 out of 36 layers reduces inference FLOPs by up to $33\%$ while maintaining comparable task success with the default model across 3,000 evaluation episodes ($46.7\%$ vs. $46.9\%$), and the same principle carries over to real-robot deployment of $\pi_{0.5}$.

Key Points

Vision-language-action (VLA) models are a promising approach to general purpose robot manipulation
While the internal computation of vision-language models, which compose the backbone of VLAs, has been studied extensively, less is known about how their perceptual and semantic representations are transformed into action in VLA models.
We study this question mechanistically in three frontier VLA models ($\pi_{0.5}$, MolmoAct2, and MolmoBot) by analyzing the causal pathways through which the input image and the instruction shape actions.
Across the three models, activation patching shows that causally relevant information influences the action expert (AE) through a narrow band of mid-to-late cross-attention layers, which we term the action-conditioning interface.
The content of this information changes with token and layer within the model

Discovering the Action-Conditioning Interface

Our goal is to study how task-relevant information within the vision-language backbone is passed to the action expert. We then use the identified causal interface to create policies that improve efficiency without costing performance. We study MolmoBot, $\pi_{0.5}$, and MolmoAct2 ( Deshpande et al., 2026 ; Physical Intelligence et al., 2025 ; Fang et al., 2026 ) .

Each model conditions its continuous action expert on representations from all backbone layers using different attention implementations.

Discovering the Action-Conditioning Interface — Causal Analysis of Robot Foundation Models Enables Efficiency Gains — Figure 1: Activation patching reveals how a vision-language-action (VLA) model turns perception into action. ( Left ) We copy activations from a donor episode (“Pick up the banana”) into a target episode (“Pick up the apple”) with the position of the objects swapped. Visual information (purple) sits in early image tokens, semantic information (blue) in the early noun token, and geometric information (orange) in later text tokens. The action expert (AE) reads from a narrow band of later layer. ( Top right ) Schematic of these effects by layer. The action expert reads from a narrow band of layers.( Bottom right ) Keeping only this interface in MolmoBot, 3 of 36 connections , cuts compute by 33% with comparable average success (46.7% vs. 46.9%) over 3000 trials.

Efficiency Improvements in Real Policies

Our mechanistic findings suggest a direct opportunity to improve the efficiency of VLA architectures: the narrow VLM–action-expert interface identified in Figure 2 . Thus, we restrict action expert conditioning to the identified layers by removing completely the cross attention layers outside of the action-conditioning interface , and evaluate whether these simplified policies retain their capabilities.

Explaining the Action Conditioning Interface

In a VLA, information flows from perception, through the vision-language backbone, to the action expert that produces motor commands. The interface discovery in Section 2 raises a question about what information the backbone supplies to the expert. Our goal is to distinguish representations that determine which object the model selects from those that determine where it reaches.

We test hypotheses about this information by comparing the effects of interchange interventions with predictions from a high-level causal model.

Explaining the Action Conditioning Interface — Causal Analysis of Robot Foundation Models Enables Efficiency Gains — (b) Predicted reaches under the labeled transfer hypotheses.

Sources and Demonstrations

Figures are reproduced from Swann et al. See the full paper for experimental details and the project page, when the paper links one, for demonstrations and videos.