# PuzzleBench: A Fully Dynamic Evaluation Framework for Large Multimodal Models on Puzzle Solving

arXiv: https://arxiv.org/abs/2504.10885 · HTML: https://mlsanity.com/paper/2504.10885/

## Key facts

- Cited by 17 papers in the map (data through 14 Oct 2026).
- Placed in VLM benchmarks and visual reasoning, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 20 of its 39 references that are in the map.
- First posted on arXiv 15 Apr 2025.

## Placed in

- Multimodal LLMs / vision-language models / [VLM benchmarks and visual reasoning](https://mlsanity.com/mllm/vlm-benchmarks-and-visual-reasoning/)

## Builds on (most cited first, 20 of 20)

- [Microsoft COCO: Common Objects in Context](https://mlsanity.com/paper/1405.0312/) (2014)
- [Training Verifiers to Solve Math Word Problems](https://mlsanity.com/paper/2110.14168/) (2021)
- [Evaluating Large Language Models Trained on Code](https://mlsanity.com/paper/2107.03374/) (2021)
- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution](https://mlsanity.com/paper/2409.12191/) (2024)
- [LLaVA-OneVision: Easy Visual Task Transfer](https://mlsanity.com/paper/2408.03326/) (2024)
- [MMBench: Is Your Multi-modal Model an All-around Player?](https://mlsanity.com/paper/2307.06281/) (2023)
- [Towards VQA Models That Can Read](https://mlsanity.com/paper/1904.08920/) (2019)
- [Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling](https://mlsanity.com/paper/2412.05271/) (2024)
- [DocVQA: A Dataset for VQA on Document Images](https://mlsanity.com/paper/2007.00398/) (2020)
- [MM-Vet: Evaluating Large Multimodal Models for Integrated Capabilities](https://mlsanity.com/paper/2308.02490/) (2023)
- [DeepSeek-VL2: Mixture-of-Experts Vision-Language Models for Advanced Multimodal Understanding](https://mlsanity.com/paper/2412.10302/) (2024)
- [LogicVista: Multimodal LLM Logical Reasoning Benchmark in Visual Contexts](https://mlsanity.com/paper/2407.04973/) (2024)
- [ChatGPT: Jack of all trades, master of none](https://mlsanity.com/paper/2302.10724/) (2023)
- [Rethinking Benchmark and Contamination for Language Models with Rephrased Samples](https://mlsanity.com/paper/2311.04850/) (2023)
- [Exploring the Reasoning Abilities of Multimodal Large Language Models (MLLMs): A Comprehensive Survey on Emerging Trends in Multimodal Reasoning](https://mlsanity.com/paper/2401.06805/) (2024)
- [Task Contamination: Language Models May Not Be Few-Shot Anymore](https://mlsanity.com/paper/2312.16337/) (2023)
- [GMAI-MMBench: A Comprehensive Multimodal Evaluation Benchmark Towards General Medical AI](https://mlsanity.com/paper/2408.03361/) (2024)
- [NPHardEval: Dynamic Benchmark on Reasoning Ability of Large Language Models via Complexity Classes](https://mlsanity.com/paper/2312.14890/) (2023)
- [OBI-Bench: Can LMMs Aid in Study of Ancient Script on Oracle Bones?](https://mlsanity.com/paper/2412.01175/) (2024)

## Built on by (most cited first, 17 of 17)

- [Thinking with Images for Multimodal Reasoning: Foundations, Methods, and Future Frontiers](https://mlsanity.com/paper/2506.23918/) (2025)
- [Perception, Reason, Think, and Plan: A Survey on Large Multimodal Reasoning Models](https://mlsanity.com/paper/2505.04921/) (2025)
- [Video-Thinker: Sparking "Thinking with Videos" via Reinforcement Learning](https://mlsanity.com/paper/2510.23473/) (2025)
- [MME-Reasoning: A Comprehensive Benchmark for Logical Reasoning in MLLMs](https://mlsanity.com/paper/2505.21327/) (2025)
- [ABench-Physics: Benchmarking Physical Reasoning in LLMs via High-Difficulty and Dynamic Physics Problems](https://mlsanity.com/paper/2507.04766/) (2025)
- [VisualSphinx: Large-Scale Synthetic Vision Logic Puzzles for RL](https://mlsanity.com/paper/2505.23977/) (2025)
- [Can Large Models Fool the Eye? A New Turing Test for Biological Animation](https://mlsanity.com/paper/2508.06072/) (2025)
- [PuzzlePlex: Benchmarking Foundation Models on Reasoning and Planning with Puzzles](https://mlsanity.com/paper/2510.06475/) (2025)
- [Pencil Puzzle Bench: A Benchmark for Multi-Step Verifiable Reasoning](https://mlsanity.com/paper/2603.02119/) (2026)
- [VRIQ: Benchmarking and Analyzing Visual-Reasoning IQ of VLMs](https://mlsanity.com/paper/2602.05382/) (2026)
- [Reasoning or Pattern Matching? Probing Large Vision-Language Models with Visual Puzzles](https://mlsanity.com/paper/2601.13705/) (2026)
- [Reasoning Structure of Large Language Models](https://mlsanity.com/paper/2606.03883/) (2026)
- [ComplexityWorld: Benchmarking Vision-Language Models on Verifiable Visual Decision Making](https://mlsanity.com/paper/2608.07584/) (2026)
- [Invisible in Space, Visible in Time: Motion Vision CAPTCHA against GUI Agents](https://mlsanity.com/paper/2609.27461/) (2026)
- [LongPuzzleBench: Evaluating GUI Agents on Long-Horizon Visual Puzzles](https://mlsanity.com/paper/2609.34769/) (2026)
- [AdaC-GRPO: Towards Mastering Long-Horizon Visual Sudoku Puzzles via Adaptive Curriculum Learning](https://mlsanity.com/paper/2610.13184/) (2026)
- [Query Your Way Out: Benchmarking Multi-Turn Knowledge Reasoning through Generalized Puzzle Solving](https://mlsanity.com/paper/2610.16133/) (2026)

## Abstract (arXiv)

Large Multimodal Models (LMMs) have demonstrated impressive capabilities across a wide range of multimodal tasks, achieving ever-increasing performance on various evaluation benchmarks. However, existing benchmarks are typically static and often overlap with pre-training datasets, leading to fixed complexity constraints and substantial data contamination issues. Meanwhile, manually annotated datasets are labor-intensive, time-consuming, and subject to human bias and inconsistency, leading to reliability and reproducibility issues. To address these problems, we propose a fully dynamic multimodal evaluation framework, named Open-ended Visual Puzzle Generation (OVPG), which aims to generate fresh, diverse, and verifiable evaluation data automatically in puzzle-solving tasks. Specifically, the OVPG pipeline consists of a raw material sampling module, a visual content generation module, and a puzzle rule design module, which ensures that each evaluation instance is primitive, highly randomized, and uniquely solvable, enabling continual adaptation to the evolving capabilities of LMMs. Built upon OVPG, we construct PuzzleBench, a dynamic and scalable benchmark comprising 11,840 VQA samples. It features six carefully designed puzzle tasks targeting three core LMM competencies, visual recognition, logical reasoning, and context understanding. PuzzleBench differs from static benchmarks that quickly become outdated. It enables ongoing dataset refreshing through OVPG and a rich set of open-ended puzzle designs, allowing seamless adaptation to the evolving capabilities of LMMs.