SpatialClaw Accepted to NeurIPS 2026: Free Code-Action Harness Lifts Frontier Models Up to 15 Points on 20 Spatial Benchmarks
CMHungSteven · x · 2026-09-29
SpatialClaw is accepted to NeurIPS 2026. The training-free, model-agnostic harness — a persistent Python kernel where the VLM writes one cell per step, inspects SAM/depth/geometry outputs, revises, then answers — lifts four frontier models across 20 spatial benchmarks:
- GPT-6 Astra: 71.3 → 77.4 (+6.1)
- Opus 5: 58.5 → 73.5 (+15.0)
- GPT-6 Sol: 61.2 → 72.9 (+11.7)
- Qwen3.8-27B: 59.8 → 67.3 (+7.5)
Notably, even GPT-6 Astra, which already looks "spatial," improves with the kernel — showing the action interface itself remains a lever. Paper, code and website are public.
More from Research
- Amazon's GEB grounds entity biographies for long-video memory, hits 72% on EgoLifeQA — amazon · 2026-09-30
- NVIDIA's LongLive-Plug: Distill Once, Deploy Training-Free Across 54 Downstream Video Models — nvidia · 2026-09-30
- CrossBFM Distills a Shared Behavior Space Across Humanoid Robots in Under One GPU-Hour — Tan-Dzung Do · 2026-09-30
- AnisoWM: anisotropic representations improve planning in JEPA world models — SeoulNatlUniv · 2026-09-30
- Meituan's SAKI: Coupling-Routed Teacher Supervision Speeds Up On-Policy Distillation 4.22x — meituan · 2026-09-30
- Real2Gym Turns Videos Into Interactive Robot Gyms, Beating GPT-6 Direct Mode by 33% on Real Robots — Shanghai-AI-Laboratory · 2026-09-30