Do VLAs Understand and Adapt to the Objects They Handle, or Simply Replay Learned Behaviors?
cs.RO, cs.AI
Submitted: 2026-10-05
Updated: 2026-10-05
Sources
- LIBERO-PRO: Towards Robust and Fair Evaluation of Vision-Language-Action Models Beyond Memorization
- LIBERO-Para: A Diagnostic Benchmark and Metrics for Paraphrase Robustness in VLA Models
- Probing a Vision-Language-Action Model for Symbolic States and Integration into a Cognitive Architecture
- What Frozen VLAs Already Know About Success: A Probing Study of Value-Like Structure in Foundation Robot Policies
- Observing and Controlling Features in Vision-Language-Action Models
- Sparse Autoencoders Reveal Interpretable and Steerable Features in VLA Models
- Enhancing Generalization in Vision-Language-Action Models by Preserving Pretrained Representations
- VLA-Trace: Diagnosing Vision-Language-Action Models through Representation and Behavior Tracing
- DROID: A Large-Scale In-The-Wild Robot Manipulation Dataset
- FAST: Efficient Action Tokenization for Vision-Language-Action Models
- OpenVLA: An Open-Source Vision-Language-Action Model
- $\pi_0$: A Vision-Language-Action Flow Model for General Robot Control
- $\pi_{0.5}$: a Vision-Language-Action Model with Open-World Generalization
- SmolVLA: A Vision-Language-Action Model for Affordable and Efficient Robotics
- GR00T N1: An Open Foundation Model for Generalist Humanoid Robots
- VLA-JEPA: Enhancing Vision-Language-Action Model with Latent World Model
- PaliGemma: A versatile 3B VLM for transfer
- SmolVLM: Redefining small and efficient multimodal models
- Qwen3-VL Technical Report
- Prismatic VLMs: Investigating the Design Space of Visually-Conditioned Language Models
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
- ARC: A Reasoning Recipe for Robot Foundation Models
- HRDexDB: A 4D Dexterous Grasping Dataset Across Human and Multiple Robot Embodiments
- APT: Action Expert Pretraining Improves Instruction Generalization of Vision-Language-Action Policies