MM-CondChain: A Programmatically Verified Benchmark for Visually Grounded Deep Compositional Reasoning

TL;DR

This paper introduces MM-CondChain, a new benchmark for evaluating deep compositional reasoning in multimodal models, emphasizing multi-layer visual conditions that require detailed perception and complex reasoning.

Contribution

It presents a scalable agentic synthesis pipeline with verifiable layers to construct challenging, multi-domain benchmarks for visual reasoning tasks.

Findings

MLLMs achieve only 53.33 Path F1 on the benchmark

Performance drops significantly on hard negatives and deeper chains

Deep compositional reasoning remains a key challenge for current models

Abstract

Multimodal Large Language Models (MLLMs) are increasingly used to carry out visual workflows such as navigating GUIs, where the next step depends on verified visual compositional conditions (e.g., "if a permission dialog appears and the color of the interface is green, click Allow") and the process may branch or terminate early. Yet this capability remains under-evaluated: existing benchmarks focus on shallow-compositions or independent-constraints rather than deeply chained compositional conditionals. In this paper, we introduce MM-CondChain, a benchmark for visually grounded deep compositional reasoning. Each benchmark instance is organized as a multi-layer reasoning chain, where every layer contains a non-trivial compositional condition grounded in visual evidence and built from multiple objects, attributes, or relations. To answer correctly, an MLLM must perceive the image in detail, reason over multiple visual elements at each step, and follow the resulting execution path to the final outcome. To scalably construct such workflow-style data, we propose an agentic synthesis pipeline: a Planner orchestrates layer-by-layer generation of compositional conditions, while a Verifiable Programmatic Intermediate Representation (VPIR) ensures each layer's condition is mechanically verifiable. A Composer then assembles these verified layers into complete instructions. Using this pipeline, we construct benchmarks across three visual domains: natural images, data charts, and GUI trajectories. Experiments on a range of MLLMs show that even the strongest model attains only 53.33 Path F1, with sharp drops on hard negatives and as depth or predicate complexity grows, confirming that deep compositional reasoning remains a fundamental challenge.