AI Paper Atlas

Research map / Multi-agent systems and games / Human-AI and goal-conditioned

PromptPilot: Game-Theoretic Multi-Agent Prompt Optimization for Segment Anything

Guangze Shi, Yingjie Mi, Jia Shen, Feixue Shao, Jiarui Cao, Yexin Lai, Xueyu Liu, Rui Wang, Yongfei Wu, Mingqiang Wei

ICML 2026 · Multi-agent systems and games › Human-AI and goal-conditioned · approach: Collaborative multi-agent perception

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 manager agent with SAM DSC feedback to feature-only agent (ablation)single feature agent · 4 medical datasets avgmIoU: 34.0 -> 52.2Improved
add manager agent with SAM DSC feedback to physical-only agent (ablation)single physical agent · 4 medical datasets avgmIoU: 34.5 -> 53.0Improved
combine feature and physical agents under the manager (ablation)DSC-manager variant · 4 medical datasets avgmIoU: 52.2 -> 61.3Improved
add LOO marginal contribution with EMA to the manager reward (ablation)PromptPilot · 4 medical datasets avgmIoU: 61.3 -> 62.9Improved
use feature objective only versus physical objective only (single agent) (ablation)single-agent prompt optimization · 4 medical datasets avgmIoU: 34.0 -> 34.5No clear effect
use one monolithic DQN agent for both objectives instead of separate agents (ablation)single-agent prompt optimization · 4 medical datasets avgmIoU: 34.2No clear effect
replace Matcher heuristic prompt selection with PromptPilotMatcher · COCOmIoU: +8.7Improved
replace single-agent PPO with multi-agent PromptPilotFM-PPO · VOC and COCOmIoU: +8.5 and +16.7Improved
replace FM-PPO with PromptPilot on feature-matching initial promptsFM-PPO · 4 medical datasets avgmIoU: 57.0 -> 62.9Improved
apply PromptPilot to coarse UNet pseudo-label promptsCoarse Segmentation baseline · 4 medical datasets avgmIoU: 25.0 -> 61.2Improved
replace PPO with PromptPilot on coarse segmentation promptsCS-PPO · 4 medical datasets avgmIoU: 28.6 -> 61.2Improved
add PromptPilot over Matcher on DAVIS 2017Matcher · DAVIS 2017region similarity J: 65.1 -> 69.3Improved

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