Research map / Vision-language and multimodal / Robot action and navigation
A Generalist Pair-wise Progress Critic Model for Vision-Language-Action Robots
Qi Zhang, Shaopeng Zhai, Shengzhe Zhang, Litao Liu, Tianyi Zhang, Fuxian Huang, Ming Zhou
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 pair-wise progress understanding pretraining (ablation) | VLAC action generation · Real-world manipulation (5 tasks) | average success rate (%): 16 -> 75 | Improved |
| use semantic-space numeric actions instead of FAST tokenizer (ablation) | VLAC · Real-world manipulation (5 tasks) | average success rate (%): 66 -> 75 | Improved |
| use VLAC instead of pi0 under lighting disturbance | VLA policy · Real-world manipulation, lighting transfer | average success rate (%): 3 -> 57 | Improved |
| VLAC instead of pi0 in-distribution | VLA policy · Real-world manipulation (5 tasks) | average success rate (%): 27 -> 75 | Improved |
| pair-wise progress prediction instead of single-point estimation (ablation) | VLAC-8b critic · eight critic datasets | VOC | Improved |
| add one reference example (in-context one-shot) (ablation) | VLAC-8b critic · RoboNet | VOC-F1: 0 -> improves significantly | Improved |
| add Ego4D human video to training (ablation) | VLAC-8b critic · EgoDex and RoboFAC | VOC-F1 | Improved |
| add large-scale VLAC pretraining before post-training (ablation) | critic post-training · EgoDex | VOC-F1: 0.24 -> 0.68 | Improved |
| add large-scale VLAC pretraining before post-training (ablation) | critic post-training · RoboFAC-fail | VOC-F1: 0.21 -> 0.37 | Improved |
| mix 10% failure trajectories into RoboFAC-success training (ablation) | VLAC · RoboFAC-success | VOC-F1: 0.99 -> 0.95 | Hurt |
| use dense progress-based reward instead of sparse reward in real-world RL | VLAC with PPO · Real-world RL | success rate: 30% -> 90% | Improved |
Closest papers by meaning
- VLA-R1: Enhancing Reasoning in Vision-Language-Action Models (ECCV 2026 · similarity 0.93)
- ElegantVLA: Learning When to Think for Efficient Vision-Language-Action Models (NeurIPS 2026 · similarity 0.92)
- NS-VLA: Towards Neuro-Symbolic Vision-Language-Action Models (NeurIPS 2026 · similarity 0.91)
- VLA-Adapter: An Effective Paradigm for Tiny-Scale Vision-Language-Action Model (AAAI 2026 · similarity 0.91)
- Beyond Data Scaling: Representation-Centric Pre-training for Vision-Language-Action Models (NeurIPS 2026 · similarity 0.90)
- VLA-REPLICA: A Low-Cost, Reproducible Benchmark for Real-World Evaluation of Vision-Language-Action Models (NeurIPS 2026 · similarity 0.90)
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.
