Paraphrase Consensus Is Not Self-Consistency: Instruction Reliance and Robustness in VLA Policies

arXiv:2610.19438 · cs.RO · Submitted 2026-10-14 · Read on arXiv

cs.RO

Submitted: 2026-10-14

Updated: 2026-10-14

Comments: 30 pages (8 pages main text)

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• Consistency — 1× in this paper · explained in 3 episodes
• Robustness — 1× in this paper · explained in 5 episodes
• Self-consistency — 1× in this paper · explained in 2 episodes

Sources

• A Simple Zero-shot Prompt Weighting Technique to Improve Prompt Ensembling in Text-Image Models
• A Gentle Introduction to Conformal Prediction and Distribution-Free Uncertainty Quantification
• It's MBR All the Way Down: Modern Generation Techniques Through the Lens of Minimum Bayes Risk
• $\pi_{0.5}$: a Vision-Language-Action Model with Open-World Generalization
• $\pi_0$: A Vision-Language-Action Flow Model for General Robot Control
• LIBERO-Plus: In-depth Robustness Analysis of Vision-Language-Action Models
• LangGap: Diagnosing and Closing the Language Gap in Vision-Language-Action Models
• Verifier-free Test-Time Sampling for Vision-Language-Action Models
• LIBERO-Para: A Diagnostic Benchmark and Metrics for Paraphrase Robustness in VLA Models
• Fine-Tuning Vision-Language-Action Models: Optimizing Speed and Success
• Perturbation-Based Uncertainty for Failure Detection in Vision-Language-Action Models
• Quantifying Language Models' Sensitivity to Spurious Features in Prompt Design or: How I learned to start worrying about prompt formatting
• VLA-FAIL: Efficient Task Failure Detection for Finetuned Vision-Language-Action Models
• Evaluating the Zero-shot Robustness of Instruction-tuned Language Models
• Unsupervised Data Augmentation for Consistency Training
• Grounded Semantic Re-Binding for Robust Instruction Generalization in Vision-Language-Action Models
• SPRINT: Scalable Policy Pre-Training via Language Instruction Relabeling
• LIBERO-PRO: Towards Robust and Fair Evaluation of Vision-Language-Action Models Beyond Memorization

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