Learning Deliberately, Acting Intuitively: Unlocking Test-Time Reasoning in Multimodal LLMs

• URL: http://arxiv.org/abs/2507.06999v1
• Date: Wed, 09 Jul 2025 16:25:44 GMT
• Title: Learning Deliberately, Acting Intuitively: Unlocking Test-Time Reasoning in Multimodal LLMs
• Authors: Yahan Yu, Yuyang Dong, Masafumi Oyamada,
• Abstract summary: Deliberate-to-Intuitive reasoning framework (D2I) improves understanding and reasoning ability of multimodal language models.<n>Our method sets deliberate reasoning strategies to enhance modality alignment only through the rule-based format reward during training.<n>While evaluating, the reasoning style shifts to intuitive, which removes deliberate reasoning strategies during training and implicitly reflects the model's acquired abilities in the response.
• Score: 7.501387372794562
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Reasoning is a key capability for large language models (LLMs), particularly when applied to complex tasks such as mathematical problem solving. However, multimodal reasoning research still requires further exploration of modality alignment and training costs. Many of these approaches rely on additional data annotation and relevant rule-based rewards to enhance the understanding and reasoning ability, which significantly increases training costs and limits scalability. To address these challenges, we propose the Deliberate-to-Intuitive reasoning framework (D2I) that improves the understanding and reasoning ability of multimodal LLMs (MLLMs) without extra annotations and complex rewards. Specifically, our method sets deliberate reasoning strategies to enhance modality alignment only through the rule-based format reward during training. While evaluating, the reasoning style shifts to intuitive, which removes deliberate reasoning strategies during training and implicitly reflects the model's acquired abilities in the response. D2I outperforms baselines across both in-domain and out-of-domain benchmarks. Our findings highlight the role of format reward in fostering transferable reasoning skills in MLLMs, and inspire directions for decoupling training-time reasoning depth from test-time response flexibility.

Related papers

• VL-Cogito: Progressive Curriculum Reinforcement Learning for Advanced Multimodal Reasoning [69.44871115752055]
  We propose an advanced multimodal reasoning model trained via a novel Progressive Curriculum Reinforcement Learning (PCuRL) framework.<n>PCuRL systematically guides the model through tasks of gradually increasing difficulty, substantially improving its reasoning abilities across diverse multimodal contexts.<n>The framework introduces two key innovations: (1) an online difficulty soft weighting mechanism, dynamically adjusting training difficulty across successive RL training stages; and (2) a dynamic length reward mechanism, which encourages the model to adaptively regulate its reasoning path length according to task complexity.
  arXiv  Detail & Related papers  (2025-07-30T12:23:21Z)
• 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)
• Perceptual Decoupling for Scalable Multi-modal Reasoning via Reward-Optimized Captioning [78.17782197231325]
  We propose a reasoning-guided reinforcement learning strategy that aligns the extractor's captioning behavior with the reasoning objective.<n> Experiments on multi-modal math and science benchmarks show that the proposed RACRO method achieves state-of-the-art average performance.
  arXiv  Detail & Related papers  (2025-06-05T02:28:07Z)
• SRPO: Enhancing Multimodal LLM Reasoning via Reflection-Aware Reinforcement Learning [25.02860760920562]
  Multimodal large language models (MLLMs) have shown promising capabilities in reasoning tasks, but struggle with complex problems requiring explicit self-reflection and self-correction.<n>Existing reflection methods are simplistic and struggle to generate meaningful and instructive feedback.<n>We propose Multimodal Self-Reflection enhanced reasoning with Group Relative Policy Optimization (SRPO), a two-stage reflection-aware reinforcement learning framework.
  arXiv  Detail & Related papers  (2025-06-02T14:21:44Z)
• Incentivizing Dual Process Thinking for Efficient Large Language Model Reasoning [75.04643265875072]
  Large reasoning models (LRMs) have demonstrated strong performance on complex reasoning tasks, but often suffer from overthinking.<n>Inspired by the dual process theory in cognitive science, we propose Adaptive Cognition Policy Optimization.<n>ACPO enables LRMs to achieve efficient reasoning through adaptive cognitive allocation and dynamic system switch.
  arXiv  Detail & Related papers  (2025-05-22T07:15:08Z)
• Learning to Reason via Mixture-of-Thought for Logical Reasoning [56.24256916896427]
  Mixture-of-Thought (MoT) is a framework that enables LLMs to reason across three complementary modalities: natural language, code, and truth-table.<n>MoT adopts a two-phase design: (1) self-evolving MoT training, which jointly learns from filtered, self-generated rationales across modalities; and (2) MoT inference, which fully leverages the synergy of three modalities to produce better predictions.
  arXiv  Detail & Related papers  (2025-05-21T17:59:54Z)
• When Thinking Fails: The Pitfalls of Reasoning for Instruction-Following in LLMs [16.659986373052217]
  Chain-of-thought reasoning can significantly degrade instruction-following accuracy.<n>This is the first work to systematically expose reasoning-induced failures in instruction-following.
  arXiv  Detail & Related papers  (2025-05-16T16:36:00Z)
• Guiding Reasoning in Small Language Models with LLM Assistance [23.3038074903744]
  Small Language Models cast doubt suitability for tasks demanding deep, multi-step logical deduction.<n>This paper introduces a framework called Small Reasons, Large Hints, which selectively augments SLM reasoning with targeted guidance from large language models.<n>Our experiments on mathematical reasoning datasets demonstrate that targeted external scaffolding significantly improves performance.
  arXiv  Detail & Related papers  (2025-04-14T06:32:45Z)
• Why Reasoning Matters? A Survey of Advancements in Multimodal Reasoning (v1) [66.51642638034822]
  Reasoning is central to human intelligence, enabling structured problem-solving across diverse tasks.<n>Recent advances in large language models (LLMs) have greatly enhanced their reasoning abilities in arithmetic, commonsense, and symbolic domains.<n>This paper offers a concise yet insightful overview of reasoning techniques in both textual and multimodal LLMs.
  arXiv  Detail & Related papers  (2025-04-04T04:04:56Z)
• DDCoT: Duty-Distinct Chain-of-Thought Prompting for Multimodal Reasoning in Language Models [28.712359821231182]
  Large language models (LLMs) have made remarkable strides in such multi-step reasoning on the language modality solely by leveraging the chain of thought (CoT) to mimic human thinking. The transfer of these advancements to multimodal contexts introduces heightened challenges, including but not limited to the impractical need for labor-intensive annotation. This study proposes a novel DDCoT prompting that maintains a critical attitude through negative-space prompting and incorporates multimodality into reasoning.
  arXiv  Detail & Related papers  (2023-10-25T08:03:10Z)
• Towards LogiGLUE: A Brief Survey and A Benchmark for Analyzing Logical Reasoning Capabilities of Language Models [56.34029644009297]
  Large language models (LLMs) have demonstrated the ability to overcome various limitations of formal Knowledge Representation (KR) systems. LLMs excel most in abductive reasoning, followed by deductive reasoning, while they are least effective at inductive reasoning. We study single-task training, multi-task training, and "chain-of-thought" knowledge distillation fine-tuning technique to assess the performance of model.
  arXiv  Detail & Related papers  (2023-10-02T01:00:50Z)

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.