Jigsaw-R1: A Study of Rule-based Visual Reinforcement Learning with Jigsaw Puzzles

• URL: http://arxiv.org/abs/2505.23590v2
• Date: Mon, 02 Jun 2025 09:10:33 GMT
• Title: Jigsaw-R1: A Study of Rule-based Visual Reinforcement Learning with Jigsaw Puzzles
• Authors: Zifu Wang, Junyi Zhu, Bo Tang, Zhiyu Li, Feiyu Xiong, Jiaqian Yu, Matthew B. Blaschko,
• Abstract summary: This paper provides a comprehensive study of rule-based visual RL, using jigsaw puzzles as a structured experimental framework.<n> Jigsaw puzzles offer inherent ground truth, adjustable difficulty, and demand complex decision-making.<n> MLLMs, initially performing near to random guessing on the simplest jigsaw puzzles, achieve near-perfect accuracy and generalize to complex, unseen configurations through fine-tuning.
• Score: 22.005722971314707
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: The application of rule-based reinforcement learning (RL) to multimodal large language models (MLLMs) introduces unique challenges and potential deviations from findings in text-only domains, particularly for perception-heavy tasks. This paper provides a comprehensive study of rule-based visual RL, using jigsaw puzzles as a structured experimental framework. Jigsaw puzzles offer inherent ground truth, adjustable difficulty, and demand complex decision-making, making them ideal for this study. Our research reveals several key findings: \textit{Firstly,} we find that MLLMs, initially performing near to random guessing on the simplest jigsaw puzzles, achieve near-perfect accuracy and generalize to complex, unseen configurations through fine-tuning. \textit{Secondly,} training on jigsaw puzzles can induce generalization to other visual tasks, with effectiveness tied to specific task configurations. \textit{Thirdly,} MLLMs can learn and generalize with or without explicit reasoning, though open-source models often favor direct answering. Consequently, even when trained for step-by-step reasoning, they can ignore the thinking process in deriving the final answer. \textit{Fourthly,} we observe that complex reasoning patterns appear to be pre-existing rather than emergent, with their frequency increasing alongside training and task difficulty. \textit{Finally,} our results demonstrate that RL exhibits more effective generalization than Supervised Fine-Tuning (SFT), and an initial SFT cold start phase can hinder subsequent RL optimization. Although these observations are based on jigsaw puzzles and may vary across other visual tasks, this research contributes a valuable piece of jigsaw to the larger puzzle of collective understanding rule-based visual RL and its potential in multimodal learning. The code is available at: https://github.com/zifuwanggg/Jigsaw-R1.

Related papers

• RL for Reasoning by Adaptively Revealing Rationales [36.50924054394857]
  Supervised fine-tuning (SFT) relies on dense ground-truth labels, which become increasingly costly as sequence length grows.<n>We address this by adaptive backtracking (AdaBack), a per-sample curriculum learning algorithm that reveals only a partial prefix of the target output during training.<n>We show that our adaptive curriculum over partial answers reliably solves problems that are otherwise intractable.
  arXiv  Detail & Related papers  (2025-06-22T17:46:14Z)
• LogicPuzzleRL: Cultivating Robust Mathematical Reasoning in LLMs via Reinforcement Learning [29.047063129464494]
  Large language models (LLMs) excel at many supervised tasks but often struggle with structured reasoning unfamiliar settings.<n>This discrepancy suggests that standard fine-tuning pipelines may instill narrow, domain-specifics rather than fostering general-purpose thinking strategies.<n>We propose a "play to learn" framework that fine-tunes LLMs through reinforcement learning on a suite of seven custom logic puzzles.
  arXiv  Detail & Related papers  (2025-06-05T09:40:47Z)
• ERL-MPP: Evolutionary Reinforcement Learning with Multi-head Puzzle Perception for Solving Large-scale Jigsaw Puzzles of Eroded Gaps [28.009783235854584]
  We propose a framework of Evolutionary Reinforcement Learning with Multi-head Puzzle Perception.<n>The proposed ERL-MPP is evaluated on the JPLEG-5 dataset with large gaps and the MIT dataset with large-scale puzzles.<n>It significantly outperforms all state-of-the-art models on both datasets.
  arXiv  Detail & Related papers  (2025-04-13T14:56:41Z)
• Fast-Slow-Thinking: Complex Task Solving with Large Language Models [49.98959729052245]
  This paper introduces a new task decomposition method termed Fast-Slow-Thinking'' (FST)<n>In FT, LLMs are prompted to remove the constraints of the original task, therefore simplifying it to a general and concise one.<n>In ST, we recall the constraints removed in FT, so that LLMs can improve the answer generated in FT to meet the requirements of the original task.
  arXiv  Detail & Related papers  (2025-04-11T16:57:36Z)
• Exploring the Effect of Reinforcement Learning on Video Understanding: Insights from SEED-Bench-R1 [53.894789613838654]
  We introduce SEED-Bench-R1, a benchmark designed to evaluate post-training methods for MLLMs in video understanding.<n>It includes intricate real-world videos and complex everyday planning tasks in the format of multiple-choice questions.<n>Using Qwen2-VL-Instruct-7B as a base model, we compare RL with supervised fine-tuning (SFT)<n>Our detailed analysis reveals that RL enhances visual perception but often produces less coherent reasoning chains.
  arXiv  Detail & Related papers  (2025-03-31T17:55:23Z)
• CrossWordBench: Evaluating the Reasoning Capabilities of LLMs and LVLMs with Controllable Puzzle Generation [53.452699232071495]
  CrossWordBench is a benchmark designed to evaluate the reasoning capabilities of Large Language Models (LLMs) and Large Vision-Language Models (LVLMs) through the medium of crossword puzzles.<n>Our evaluation reveals that reasoning LLMs outperform non-reasoning models substantially by effectively leveraging crossing-letter constraints.<n>Our findings offer insights into the limitations of the reasoning capabilities of current LLMs and LVLMs, and provide an effective approach for creating multimodal constrained tasks for future evaluations.
  arXiv  Detail & Related papers  (2025-03-30T20:03:36Z)
• VGRP-Bench: Visual Grid Reasoning Puzzle Benchmark for Large Vision-Language Models [31.645103181716678]
  Large Vision-Language Models (LVLMs) struggle with puzzles, which require precise perception, rule comprehension, and logical reasoning.<n>We introduce VGRP-Bench, a Visual Grid Reasoning Puzzle Benchmark featuring 20 diverse puzzles.<n>Our results reveal that even the state-of-the-art LVLMs struggle with these puzzles, highlighting fundamental limitations in their puzzle-solving capabilities.
  arXiv  Detail & Related papers  (2025-03-29T12:50:38Z)
• OpenVLThinker: An Early Exploration to Complex Vision-Language Reasoning via Iterative Self-Improvement [91.88062410741833]
  This study investigates whether similar reasoning capabilities can be successfully integrated into large vision-language models (LVLMs)<n>We consider an approach that iteratively leverages supervised fine-tuning (SFT) on lightweight training data and Reinforcement Learning (RL) to further improve model generalization.<n>OpenVLThinker, a LVLM exhibiting consistently improved reasoning performance on challenging benchmarks such as MathVista, MathVerse, and MathVision, demonstrates the potential of our strategy for robust vision-language reasoning.
  arXiv  Detail & Related papers  (2025-03-21T17:52:43Z)
• R1-Searcher: Incentivizing the Search Capability in LLMs via Reinforcement Learning [87.30285670315334]
  textbfR1-Searcher is a novel two-stage outcome-based RL approach designed to enhance the search capabilities of Large Language Models.<n>Our framework relies exclusively on RL, without requiring process rewards or distillation for a cold start.<n>Our experiments demonstrate that our method significantly outperforms previous strong RAG methods, even when compared to the closed-source GPT-4o-mini.
  arXiv  Detail & Related papers  (2025-03-07T17:14:44Z)
• TextGames: Learning to Self-Play Text-Based Puzzle Games via Language Model Reasoning [26.680686158061192]
  Reasoning is a fundamental capability of large language models (LLMs)<n>This paper introduces TextGames, a benchmark specifically crafted to assess LLMs through demanding text-based games.<n>Our findings reveal that although LLMs exhibit proficiency in addressing most easy and medium-level problems, they face significant challenges with more difficult tasks.
  arXiv  Detail & Related papers  (2025-02-25T18:26:48Z)
• Beyond Interpolation: Extrapolative Reasoning with Reinforcement Learning and Graph Neural Networks [18.982541044390384]
  This study focuses on the impact of the inductive bias of the architecture, different reward systems and the role of recurrent modeling in enabling sequential reasoning.<n>We show how these elements contribute to successful extrapolation on increasingly complex puzzles.
  arXiv  Detail & Related papers  (2025-02-06T08:07:35Z)
• Satori: Reinforcement Learning with Chain-of-Action-Thought Enhances LLM Reasoning via Autoregressive Search [57.28671084993782]
  Large language models (LLMs) have demonstrated remarkable reasoning capabilities across diverse domains.<n>Recent studies have shown that increasing test-time computation enhances LLMs' reasoning capabilities.<n>We propose a two-stage training paradigm: 1) a small-scale format tuning stage to internalize the COAT reasoning format and 2) a large-scale self-improvement stage leveraging reinforcement learning.
  arXiv  Detail & Related papers  (2025-02-04T17:26:58Z)
• Sliding Puzzles Gym: A Scalable Benchmark for State Representation in Visual Reinforcement Learning [3.8309622155866583]
  We introduce the Sliding Puzzles Gym (SPGym), a novel benchmark that reimagines the classic 8-tile puzzle with a visual observation space of images sourced from arbitrarily large datasets.<n>SPGym provides precise control over representation complexity through visual diversity, allowing researchers to systematically scale the representation learning challenge.<n>As we increase visual diversity by expanding the pool of possible images, all tested algorithms show significant performance degradation.
  arXiv  Detail & Related papers  (2024-10-17T21:23:03Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.