AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

Metric-Bench: Exploring In-context Spatial Metric Reasoning in VLMs for Indoor Scenes

Yuling Xi, Haokai Zhang, Muzhi Zhu, Hao Zhong, Zongze Du, Hengyu Zhao, Chenchen Jing, Yufei Yin, Bin Qin, Yongjie Yang, Zhenbo Luo, Hao Chen and 1 more

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Spatial reasoning in VLMs · 1 citations (OpenAlex)

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 changedWhereMeasuredEffect
add reference metrics (N_ref 0 -> 4) to the prompt (ablation)MetricReasoner · Metric-BenchMRA: 47.56 -> 48.59; RMSE 0.5540 -> 0.5323Improved
increase references from 4 to 5 (ablation)MetricReasoner · Metric-BenchMRA: 48.59 -> 46.85Hurt
add EP and BP accuracy rewards versus none (ablation)MetricReasoner · Metric-BenchMRA: 12.38 -> 48.59Improved
add BP reward on top of EP (ablation)MetricReasoner · Metric-BenchMRA: 45.95 -> 48.59Improved
add EP reward on top of BP (ablation)MetricReasoner · Metric-BenchMRA: 48.05 -> 48.59Improved
use finer BP bins (5 intervals instead of 4) (ablation)MetricReasoner · Metric-BenchMRA: 48.59 -> 46.14Hurt
use finer BP bins (6 intervals instead of 4) (ablation)MetricReasoner · Metric-BenchMRA: 48.59 -> 48.88No clear effect
use finer BP bins (8 intervals instead of 4) (ablation)MetricReasoner · Metric-BenchMRA: 48.59 -> 47.69Hurt
train with RL instead of SFT (ablation)Qwen3-VL-8B · BLINKaccuracy: 51.26% -> 81.82%Improved
train with RL instead of SFT (ablation)Qwen3-VL-8B · RoboSpatialaccuracy: 40.86 -> 55.14Improved
add CoT SFT stage before RL (ablation)MetricReasoner · BLINKaccuracy: 81.82% -> 34.53%Hurt
metric-guided RFT versus base modelQwen3-VL-8B · V* Benchaccuracy: 61.78% -> 68.59%Improved

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy