Agent Explorative Policy Optimization for Multimodal Agentic Reasoning

• URL: http://arxiv.org/abs/2605.28774v1
• Date: Wed, 27 May 2026 17:36:39 GMT
• Title: Agent Explorative Policy Optimization for Multimodal Agentic Reasoning
• Abstract summary: Agentic reasoning interleaves two behaviors with a structural asymmetry: thinking and tool use.<n>Under standard RL recipes like GRPO, the gap manifests as two diagnostic symptoms during training.<n>We propose AXPO (Agent eXplorative Policy Optimization) for each all-wrong tool-using subgroup.
• Score: 97.64835302176056
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Vision-language models with extended reasoning succeed on complex problems, but many real-world problems require external tools that internal reasoning alone often cannot resolve. Agentic reasoning therefore interleaves two behaviors with a structural asymmetry: thinking (the self-contained default) and tool use (a high-variance auxiliary acting). We refer to this asymmetry as the Thinking-Acting Gap. Under standard RL recipes like GRPO, the gap manifests as two diagnostic symptoms during training: tool use is attempted on only ~30% of rollouts, and when attempted, the tool-using rollouts within a group are all-wrong on ~40% of questions, suppressing the learning signal at the tool calls that needed it. We propose AXPO (Agent eXplorative Policy Optimization): for each all-wrong tool-using subgroup, AXPO fixes the thinking prefix and resamples the tool call and its continuation, paired with uncertainty-based prefix selection. Across nine multimodal benchmarks and three scales of Qwen3-VL-Thinking, SFT+AXPO outperforms SFT+GRPO at average (+1.8pp Pass@1 and +1.8pp Pass@4 at 8B on average) and 8B with SFT+AXPO surpasses the 32B Base on Pass@4 with 4 times fewer parameters.

Related papers

• Joint Optimization of Tool Creation and Use for Large Language Model Agents [69.07066297088187]
  SMITH (-grounded Multi-task Iterative Tool Honing) trains tool creation and tool use inside a single policy.<n>A 4B Qwen3 trained with SMITH on 13 procedural reasoning tasks with exact verifiers reaches 79.8 macro-average accuracy on held-out tasks.<n>It also reaches 40.4 on TabMWP-Hard and 42.6 on out-of-domain GQA (+7.6 over the best same-backbone inference-time baseline)
  arXiv  Detail & Related papers  (2026-08-25T13:59:34Z)
• Branching Policy Optimization: Sandbox-Native Language Agent Reinforcement Learning [14.640786318359709]
  Branching Policy Optimization (BPO) is a sandbox-native RL algorithm that snapshots the sandbox at high-entropy decision points along a backbone trajectory.<n>We prove this estimator is unbiased and has strictly lower variance than the trajectory-level baseline.<n>BPO improves success by 3.6--6.1 absolute points over GRPO and RLOO at matched compute, halves gradient-norm variance, and matches the best baseline using 38% fewer policy updates.
  arXiv  Detail & Related papers  (2026-07-15T09:37:36Z)
• ASAP: Agent-System Co-Design for Wall-Clock-Centered Auto HPO Research for ML Experiments [45.577182866886126]
  We present ASAP, an agent-system co-design for HPO.<n>We show how ASAP integrates a diverse pool of inductive-biased generalizations and to select among their proposals each round.<n>We also show how ASAP re-architects the loop to reduce end-to-end wall-clock while preserving regret quality.
  arXiv  Detail & Related papers  (2026-06-23T22:00:26Z)
• ParaVT: Taming the Tool Prior Paradox for Parallel Tool Use in Agentic Video Reinforcement Learning [91.51460129144654]
  We introduce ParaVT, the first multi-agent end-to-end RL-trained framework for Parallel Video Tool calling.<n>We show that ParaVT improves over the Qwen3-VL baseline by +7.9% on average, with PARA-GRPO lifting training-time format compliance from 0.13 to 0.64.
  arXiv  Detail & Related papers  (2026-05-19T18:01:26Z)
• MAXS: Meta-Adaptive Exploration with LLM Agents [48.04723638253802]
  MaxS is a meta-adaptive reasoning framework based on Large Language Model (LLM) Agents.<n>MAXS employs a lookahead strategy to extend reasoning paths a few steps ahead.<n>It combines step consistency variance and inter-step trend slopes to jointly select stable, consistent, and high-value reasoning steps.
  arXiv  Detail & Related papers  (2026-01-14T07:48:00Z)
• A$^2$FM: An Adaptive Agent Foundation Model for Tool-Aware Hybrid Reasoning [40.6234318894435]
  Large language models split into two families: reasoning-centric LLMs and agentic LLMs.<n>This divide arises from fundamentally different training objectives, leading to mismatched strengths and inefficiency on simple queries.<n>We present Adaptive Agent Foundation Model (A$2$FM), a unified framework that follows a route-then-align principle.
  arXiv  Detail & Related papers  (2025-10-13T17:08:25Z)
• Training Long-Context, Multi-Turn Software Engineering Agents with Reinforcement Learning [29.605396813225386]
  We show how reinforcement learning can be used to train agents for multi-turn interactive tasks.<n>Our methodology offers a practical approach for training capable agents for multi-turn interactive tasks using open-weight models.
  arXiv  Detail & Related papers  (2025-08-05T14:30:47Z)
• Adaptive Thinking via Mode Policy Optimization for Social Language Agents [75.3092060637826]
  We propose a framework to improve the adaptive thinking ability of language agents in dynamic social interactions.<n>Our framework advances existing research in three key aspects: (1) Multi-granular thinking mode design, (2) Context-aware mode switching across social interaction, and (3) Token-efficient reasoning via depth-adaptive processing.
  arXiv  Detail & Related papers  (2025-05-04T15:39:58Z)
• Acting Less is Reasoning More! Teaching Model to Act Efficiently [87.28134636548705]
  Tool-integrated reasoning augments large language models with the ability to invoke external tools to solve tasks.<n>Current approaches typically optimize only for final correctness without considering the efficiency or necessity of external tool use.<n>We propose a framework that encourages models to produce accurate answers with minimal tool calls.<n>Our approach reduces tool calls by up to 68.3% and improves tool productivity by up to 215.4%, while maintaining comparable answer accuracy.
  arXiv  Detail & Related papers  (2025-04-21T05:40:05Z)

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.