Natural State-Prediction Accuracy can Hide Weak Controlled Responsiveness in VLA Readouts

arXiv:2609.34684 · cs.RO, cs.CV, cs.LG · Submitted 2026-09-28 · Read on arXiv

cs.RO, cs.CV, cs.LG

Submitted: 2026-09-28

Updated: 2026-09-28

Terminology

• Accuracy — 33× in this paper · explained in 1 episode
• Consistency — 7× in this paper · explained in 3 episodes
• Generalization — 4× in this paper · explained in 13 episodes
• Robustness — 4× in this paper · explained in 5 episodes
• Distribution Shift — 3× in this paper · explained in 5 episodes
• Benchmark — 2× in this paper · explained in 2 episodes
• Feature Extraction — 2× in this paper · explained in 2 episodes
• Interpretability — 2× in this paper · explained in 15 episodes
• Trace — 2× in this paper · explained in 2 episodes
• VLA Models — 2× in this paper · explained in 3 episodes
• Aggregation — 1× in this paper · explained in 1 episode
• Benchmarking — 1× in this paper · explained in 3 episodes
• Calibration — 1× in this paper · explained in 7 episodes
• Causal Understanding — 1× in this paper · explained in 2 episodes
• Embeddings — 1× in this paper · explained in 2 episodes
• Failure Modes — 1× in this paper · explained in 3 episodes
• Feature Selection — 1× in this paper · explained in 4 episodes
• Regularization — 1× in this paper · explained in 3 episodes
• Semantic Understanding — 1× in this paper · explained in 2 episodes
• Steering — 1× in this paper · explained in 2 episodes
• Tokenization — 1× in this paper · explained in 2 episodes
• Verification — 1× in this paper · explained in 2 episodes

Sources

• Decoding Task Progress from VLA Representations
• $\pi_0$: A Vision-Language-Action Flow Model for General Robot Control
• Observing and Controlling Features in Vision-Language-Action Models
• Not All Features Are Created Equal: A Mechanistic Study of Vision-Language-Action Models
• OpenVLA: An Open-Source Vision-Language-Action Model
• Probing a Vision-Language-Action Model for Symbolic States and Integration into a Cognitive Architecture
• Where are we in the search for an Artificial Visual Cortex for Embodied Intelligence?
• Mechanistic Finetuning of Vision-Language-Action Models via Few-Shot Demonstrations
• Emergent World Representations in OpenVLA
• VLA-Trace: Diagnosing Vision-Language-Action Models through Representation and Behavior Tracing
• SmolVLA: A Vision-Language-Action Model for Affordable and Efficient Robotics
• DINOv3
• Sparse Autoencoders Reveal Interpretable and Steerable Features in VLA Models
• SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features
• ProbeAct: Probe-Guided Training-Free Failure Recovery in Vision-Language-Action Models
• Embodied Interpretability: Linking Causal Understanding to Generalization in Vision-Language-Action Models
• What Frozen VLAs Already Know About Success: A Probing Study of Value-Like Structure in Foundation Robot Policies

Related papers

• FMT x: An Efficient and Asymptotically Optimal Extension of the Fast Marching Tree for Dynamic Replanning
• MPCFormer: A physics-informed data-driven approach for explainable socially-aware autonomous driving
• RoboLab: A High-Fidelity Simulation Benchmark for Analysis of Task Generalist Policies
• HRDexDB: A 4D Dexterous Grasping Dataset Across Human and Multiple Robot Embodiments
• APT: Action Expert Pretraining Improves Instruction Generalization of Vision-Language-Action Policies
• Fine-tuning is Not Enough: A Parallel Framework for Collaborative Imitation and Reinforcement Learning in End-to-end Autonomous Driving