Dynamic resolution helps improve multimodal AI debates accuracy

DREAM: Dynamic Resolution Assignment For Multimodal Multi-agent Debate

Artificial IntelligenceComputation and LanguageComputer Vision and Pattern Recognition

Summary

It can be hard for AI models to look at pictures and text together and make good decisions because they usually see the same visual details, even when different parts need more or less focus. The authors designed a method where AI agents first try out different image resolutions to find the best detail level for each agent to work with. They also have a way to prevent agents from copying each other’s mistakes by tracking how sure each agent is about their answers over time. This approach made the AI teams more accurate without needing special tuning for each dataset.

What this means in practice

  • For multimodal ai developers: Enable AI systems combining images and language to selectively adjust image detail per agent for better reasoning accuracy in diverse tasks.
  • For ai model trainers: Help training teams reduce errors caused by agent conformity in ensemble models by restoring early confident correct answers.

Authors

Khanh-Binh Nguyen, Van Dai Do, Tien Anh Nguyen, Svetha Venkatesh, Hung Le

Abstract

Multi-agent debate (MAD) has emerged as an effective paradigm to improve the reasoning capabilities of large language models (LLMs) and is increasingly being extended to multimodal settings. However, existing multimodal MAD frameworks typically expose agents to the same fixed visual input, ignoring substantial variation in the visual scale needed across samples and agents. In addition, these frameworks frequently suffer from groupthink, a phenomenon where agents prematurely abandon correct deductions to conform with confident but hallucinated peer responses. To address these bottlenecks, we introduce DREAM (Dynamic Resolution Assignment For Multimodal Multi-Agent Debate), which operates via two core components: (1) Dynamic Resolution Assignment, a zero-shot probe round where agents test multiple resolutions, quantify uncertainty using Average Normalized Log-Likelihood (ANLL), and use an adaptive threshold to assign each agent to its empirically optimal resolution; (2) Uncertainty-Guided Rollback Aggregation counters groupthink by tracking each agent's uncertainty over rounds and restoring early low-uncertainty answers overridden by group pressure. On six multimodal datasets, DREAM improves the accuracy-token trade-off over multi-agent debate baselines by 1.5-3.2% accuracy without dataset-specific tuning.