SMMBench

SMMBench is a benchmark that contains 1,877 samples grounded in 264 sources and is used to evaluate agents' abilities in cross-source multimodal reasoning, conflict resolution, preference reasoning, and memory-grounded action prediction.

Papers using SMMBench (3)

SMMBench: A Benchmark for Source-Distributed Multimodal Agent Memory2026Why Keep Your Doubts to Yourself? Trading Visual Uncertainties in Multi-Agent Bandit Systems2026GAM-Agent: Game-Theoretic and Uncertainty-Aware Collaboration for Complex Visual Reasoning2025