# AMIGO: Agentic Multi-Image Grounding Oracle Benchmark

arXiv: https://arxiv.org/abs/2603.28662 · HTML: https://mlsanity.com/paper/2603.28662/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in VLM benchmarks and visual reasoning, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 22 of its 37 references that are in the map.
- First posted on arXiv 30 Mar 2026.

## Placed in

- Multimodal LLMs / vision-language models / [VLM benchmarks and visual reasoning](https://mlsanity.com/mllm/vlm-benchmarks-and-visual-reasoning/)
- LLM agents and tool use / [Multimodal and embodied agents](https://mlsanity.com/agents/multimodal-and-embodied-agents/)

## Builds on (most cited first, 22 of 22)

- [Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering](https://mlsanity.com/paper/1612.00837/) (2016)
- [MMBench: Is Your Multi-modal Model an All-around Player?](https://mlsanity.com/paper/2307.06281/) (2023)
- [MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGI](https://mlsanity.com/paper/2311.16502/) (2023)
- [MMMU-Pro: A More Robust Multi-discipline Multimodal Understanding Benchmark](https://mlsanity.com/paper/2409.02813/) (2024)
- [Kimi K2.5: Visual Agentic Intelligence](https://mlsanity.com/paper/2602.02276/) (2026)
- [GLM-4.5V and GLM-4.1V-Thinking: Towards Versatile Multimodal Reasoning with Scalable Reinforcement Learning](https://mlsanity.com/paper/2507.01006/) (2025)
- [InfographicVQA](https://mlsanity.com/paper/2104.12756/) (2021)
- [MME-RealWorld: Could Your Multimodal LLM Challenge High-Resolution Real-World Scenarios that are Difficult for Humans?](https://mlsanity.com/paper/2408.13257/) (2024)
- [mPLUG-Owl3: Towards Long Image-Sequence Understanding in Multi-Modal Large Language Models](https://mlsanity.com/paper/2408.04840/) (2024)
- [MANTIS: Interleaved Multi-Image Instruction Tuning](https://mlsanity.com/paper/2405.01483/) (2024)
- [MuirBench: A Comprehensive Benchmark for Robust Multi-image Understanding](https://mlsanity.com/paper/2406.09411/) (2024)
- [Magma: A Foundation Model for Multimodal AI Agents](https://mlsanity.com/paper/2502.13130/) (2025)
- [Can MLLMs Reason in Multimodality? EMMA: An Enhanced MultiModal ReAsoning Benchmark](https://mlsanity.com/paper/2501.05444/) (2025)
- [MMIU: Multimodal Multi-image Understanding for Evaluating Large Vision-Language Models](https://mlsanity.com/paper/2408.02718/) (2024)
- [MLLM-CompBench: A Comparative Reasoning Benchmark for Multimodal LLMs](https://mlsanity.com/paper/2407.16837/) (2024)
- [MMCR: Advancing Visual Language Model in Multimodal Multi-Turn Contextual Reasoning](https://mlsanity.com/paper/2503.18533/) (2025)
- [MIRAGE: A Benchmark for Multimodal Information-Seeking and Reasoning in Agricultural Expert-Guided Conversations](https://mlsanity.com/paper/2506.20100/) (2025)
- [MultiVerse: A Multi-Turn Conversation Benchmark for Evaluating Large Vision and Language Models](https://mlsanity.com/paper/2510.16641/) (2025)
- [MMMT-IF: A Challenging Multimodal Multi-Turn Instruction Following Benchmark](https://mlsanity.com/paper/2409.18216/) (2024)
- [InfoQuest: Evaluating Multi-Turn Dialogue Agents for Open-Ended Conversations with Hidden Context](https://mlsanity.com/paper/2502.12257/) (2025)
- [Multi-Turn Multi-Modal Question Clarification for Enhanced Conversational Understanding](https://mlsanity.com/paper/2502.11442/) (2025)
- [M3Searcher: Modular Multimodal Information Seeking Agency with Retrieval-Oriented Reasoning](https://mlsanity.com/paper/2601.09278/) (2026)

## Abstract (arXiv)

Agentic vision-language models increasingly act through extended interactions, but most evaluations still focus on single-image, single-turn correctness. We introduce AMIGO (Agentic Multi-Image Grounding Oracle Benchmark), a long-horizon benchmark for hidden-target identification over galleries of visually similar images. In AMIGO, the oracle privately selects a target image, and the model must recover it by asking a sequence of attribute-focused Yes/No questions under a strict protocol that returns Yes/No/Unsure feedback and penalizes invalid actions with Skip. This setting stresses (i) question selection under uncertainty, (ii) consistent constraint tracking across turns, and (iii) fine-grained discrimination as evidence accumulates. We instantiate AMIGO with the textitGuess My Preferred Dress task and evaluate open-source VLMs with metrics covering identification success, evidence verification, efficiency, protocol compliance, robustness to controlled feedback perturbations, and trajectory-level diagnostics. The benchmarking results show that final-answer accuracy alone overstates evidence-grounded performance: models can guess correctly without verification-passing evidence, waste turns through invalid questions, or fail to preserve the upload protocol. Strong AMIGO performance depends on the combination of visual discrimination, informative question selection, constraint tracking, efficient stopping, sustained protocol following, and recovery from controlled feedback noise; model scale alone does not guarantee reliable long-horizon interactive grounding.