Research map / Vision-language and multimodal / Robot action and navigation
NeurVLA: Unleashing Failure-Handling Capability of Vision-Language-Action Models via Neural-Symbolic Reasoning
Xuqi Liu, Minghe Gao, Juncheng Li, Siliang Tang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add observation-grounded failure correction (COG) (ablation) | InstructVLA · SimplerEnv-WidowX | success rate avg (%): 44.9 -> 55.4 | Improved |
| add trajectory-modeling failure prevention (PTM) (ablation) | InstructVLA · SimplerEnv-WidowX | success rate avg (%): 44.9 -> 52.8 | Improved |
| combine COG and PTM (ablation) | InstructVLA · SimplerEnv-WidowX | success rate avg (%): 44.9 -> 64.7 | Improved |
| replace SFT with reasoning-guided action learning (LRG) on top of COG (ablation) | InstructVLA · SimplerEnv-WidowX | success rate avg (%): 55.4 -> 62.5 | Improved |
| replace SFT with LRG on top of PTM (ablation) | InstructVLA · SimplerEnv-WidowX | success rate avg (%): 52.8 -> 59.1 | Improved |
| add LRG to COG + PTM (full NeurVLA) (ablation) | InstructVLA · SimplerEnv-WidowX | success rate avg (%): 64.7 -> 72.8 | Improved |
| use program-derived instead of LLM-direct correction ECoTs (ablation) | InstructVLA with NeurVLA · SimplerEnv-WidowX | success rate avg (%): 57.2 -> 72.8 | Improved |
| use program-derived instead of LLM-direct prevention ECoTs (ablation) | InstructVLA with NeurVLA · SimplerEnv-WidowX | success rate avg (%): 62.5 -> 72.8 | Improved |
| use program-derived instead of LLM-direct ECoTs for both (ablation) | InstructVLA with NeurVLA · SimplerEnv-WidowX | success rate avg (%): 50.4 -> 72.8 | Improved |
| apply NeurVLA to NORA | NORA · LIBERO | avg success rate (%): 87.9 -> 96.1 | Improved |
| apply NeurVLA to InstructVLA | InstructVLA · SimplerEnv | avg success rate (%): 62.3 -> 83.8 | Improved |
| apply NeurVLA to NORA | NORA · CALVIN ABC->D | avg completed length: 3.60 -> 4.09 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- ThinkAct — ThinkAct: Vision-Language-Action Reasoning via Reinforced Visual Latent Planning (NeurIPS 2025) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Failing Forward: Adaptive Failure-Informed Learning for Vision-Language-Action Models (NeurIPS 2026 · similarity 0.91)
- AHA: A Vision-Language-Model for Detecting and Reasoning Over Failures in Robotic Manipulation (ICLR 2025 · similarity 0.91)
- SAFE: Multitask Failure Detection for Vision-Language-Action Models (NeurIPS 2025 · similarity 0.90)
- NS-VLA: Towards Neuro-Symbolic Vision-Language-Action Models (NeurIPS 2026 · similarity 0.90)
- ElegantVLA: Learning When to Think for Efficient Vision-Language-Action Models (NeurIPS 2026 · similarity 0.89)
- CoT-VLA: Visual Chain-of-Thought Reasoning for Vision-Language-Action Models (CVPR 2025 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
