RoTA-Bench: VLMs show large gaps vs humans in interactive 3D projection alignment
MAC-AutoML · hf · 2026-10-16
A Xiamen University and NTU team released RoTA-Bench, which evaluates 3D spatial reasoning in vision-language models through interactive 3D-to-2D projection alignment.
- Given a target silhouette, the model observes a rotated object and must choose successive rotations to recover the projection within a limited action budget, requiring geometry inference, rotation-effect prediction and visual-feedback refinement.
- An automated pipeline constructs silhouette targets at scale for multi-turn interactive evaluation.
- Key finding: current VLMs struggle to convert 2D observations into precise 3D rotational control, showing large gaps versus humans.
More from Research
- Waypoint-OPD: Alibaba Gates On-Policy Distillation via Early-Late Teacher Shift Agreement — Bingxi Hou · 2026-10-16
- MLExpert: 35B MoE Agent Trained to Autonomously Handle Machine Learning Engineering — Shangheng Du · 2026-10-16
- VLA-Δ: Zhejiang University Adds Physical State Transition Supervision to VLA Models — OmniAI-ZJU · 2026-10-16
- Cowork-Gym: Alibaba Synthesizes 64.9K Agent Work Environments, Boosting Qwen by 8.7 Points — Anni Zou · 2026-10-16
- MJev-Spatial: Single-Pass Classification Beats Autoregressive VLMs on Spatial Reasoning — mindlab-research · 2026-10-16
- MMSearch-OPD: Joint RL and On-Policy Distillation for Long-Horizon Multimodal Search Agents — PekingUniversity · 2026-10-16