AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

SpatiO: Adaptive Test-Time Orchestration of Vision-Language Agents for Spatial Reasoning

ChanYeong Hwang, Miso Choi, Sunghyun On, Jinkyu Kim, Jungbeom Lee

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Spatial reasoning in VLMs · 0 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 the underperformance penalty delta_i in the reward (ablation)SpatiO · CV-Benchaccuracy (%): 72.0 -> 87.4Improved
use Final Reasoning Agent instead of majority voting (ablation)SpatiO · CV-Benchaccuracy (%): 77.2 -> 87.4Improved
add TTO reliability weights instead of uniform weights (ablation)SpatiO · CV-Benchaccuracy (%): 75.9 -> 87.4Improved
add short-term EMA (ablation)SpatiO · CV-Benchaccuracy (%): 83.2 -> 87.4Improved
add long-term EMA (ablation)SpatiO · CV-Benchaccuracy (%): 81.5 -> 87.4Improved
add Bayesian trust update (step 3) (ablation)SpatiO · CV-Bench 500 samplesaccuracy (%): 79.66 -> 88.12Improved
add reward computation only (step 1) (ablation)SpatiO · CV-Bench 500 samplesaccuracy (%): 86.10 -> 73.02Mixed
use trust-based role assignment instead of fixed (ablation)SpatiO · 3DSRBenchaccuracy (%): 64.8 -> 73.3Improved
use Top-k=4 instead of 3 agents (ablation)SpatiO · 3DSRBenchaccuracy (%): 73.3 -> 75.1Improved
use Top-k=5 instead of 3 agents (ablation)SpatiO · 3DSRBenchaccuracy (%): 73.3 -> 69.7Hurt
replace human role prompts with LLM-generated ones (ablation)SpatiO · 3DSRBenchaccuracy (%): 73.3 -> 71.4No clear effect
swap 2 of 5 pool agents for alternatives (ablation)SpatiO · 3DSRBenchaccuracy (%): 73.3 -> 70.9No clear effect

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