Reflecting with Two Voices: A Co-Adaptive Dual-Strategy Framework for LLM-Based Agent Decision Making

• URL: http://arxiv.org/abs/2512.08366v1
• Date: Tue, 09 Dec 2025 08:44:59 GMT
• Title: Reflecting with Two Voices: A Co-Adaptive Dual-Strategy Framework for LLM-Based Agent Decision Making
• Authors: Wentao Zhang, Qunbo Wang, Tao Zhang, Junsheng Wu, Hongping Gan, Yang Liu, Ling Dai, Shizhuang Deng, Shuntong Sun,
• Abstract summary: Large language model (LLM) agents often rely on external demonstrations or retrieval-augmented planning.<n>We propose DuSAR - a demonstration-free framework that enables a single frozen LLM to perform co-adaptive reasoning.<n>On ALFWorld and Mind2Web, DuSAR achieves state-of-the-art performance with open-source LLMs.
• Score: 24.534365665776672
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large language model (LLM) agents often rely on external demonstrations or retrieval-augmented planning, leading to brittleness, poor generalization, and high computational overhead. Inspired by human problem-solving, we propose DuSAR (Dual-Strategy Agent with Reflecting) - a demonstration-free framework that enables a single frozen LLM to perform co-adaptive reasoning via two complementary strategies: a high-level holistic plan and a context-grounded local policy. These strategies interact through a lightweight reflection mechanism, where the agent continuously assesses progress via a Strategy Fitness Score and dynamically revises its global plan when stuck or refines it upon meaningful advancement, mimicking human metacognitive behavior. On ALFWorld and Mind2Web, DuSAR achieves state-of-the-art performance with open-source LLMs (7B-70B), reaching 37.1% success on ALFWorld (Llama3.1-70B) - more than doubling the best prior result (13.0%) - and 4.02% on Mind2Web, also more than doubling the strongest baseline. Remarkably, it reduces per-step token consumption by 3-9X while maintaining strong performance. Ablation studies confirm the necessity of dual-strategy coordination. Moreover, optional integration of expert demonstrations further boosts results, highlighting DuSAR's flexibility and compatibility with external knowledge.

Related papers

• Think Fast and Slow: Step-Level Cognitive Depth Adaptation for LLM Agents [49.119608399413806]
  Large language models (LLMs) are increasingly deployed as autonomous agents for multi-turn decision-making tasks.<n>This paper introduces Cog, a framework that trains agents to dynamically adapt cognitive depth at each step.<n> Experiments on ALFWorld and ScienceWorld demonstrate that Cog achieves state-of-the-art performance with superior efficiency.
  arXiv  Detail & Related papers  (2026-02-13T06:52:09Z)
• Adversarial Attack-Defense Co-Evolution for LLM Safety Alignment via Tree-Group Dual-Aware Search and Optimization [51.12422886183246]
  Large Language Models (LLMs) have developed rapidly in web services, delivering unprecedented capabilities while amplifying societal risks.<n>Existing works tend to focus on either isolated jailbreak attacks or static defenses, neglecting the dynamic interplay between evolving threats and safeguards in real-world web contexts.<n>We propose ACE-Safety, a novel framework that jointly optimize attack and defense models by seamlessly integrating two key innovative procedures.
  arXiv  Detail & Related papers  (2025-11-24T15:23:41Z)
• ReflexGrad: Three-Way Synergistic Architecture for Zero-Shot Generalization in LLM Agents [0.0]
  We introduce ReflexGrad, a novel architecture that tightly couples three complementary mechanisms.<n>Our system achieves true zero-shot generalization through pure semantic reasoning.<n>Our work demonstrates that synergistic integration of complementary learning mechanisms enables robust zero-shot generalization.
  arXiv  Detail & Related papers  (2025-11-18T15:25:05Z)
• Graph-Enhanced Policy Optimization in LLM Agent Training [3.177432419321498]
  Group based reinforcement learning (RL) has shown impressive results on complex reasoning and mathematical tasks.<n>Group based reinforcement learning (RL) has shown impressive results on complex reasoning and mathematical tasks.
  arXiv  Detail & Related papers  (2025-10-30T08:53:41Z)
• Reinforced Strategy Optimization for Conversational Recommender Systems via Network-of-Experts [63.412646471177645]
  We propose a novel Reinforced Strategy Optimization (RSO) method for Conversational Recommender Systems (CRSs)<n>RSO decomposes the process of generating strategy-driven response decisions into the macro-level strategy planning and micro-level strategy adaptation.<n>Experiments show that RSO significantly improves interaction performance compared to state-of-the-art baselines.
  arXiv  Detail & Related papers  (2025-09-30T11:12:01Z)
• Emergent Hierarchical Reasoning in LLMs through Reinforcement Learning [56.496001894673235]
  Reinforcement Learning (RL) has proven highly effective at enhancing the complex reasoning abilities of Large Language Models (LLMs)<n>Our analysis reveals that puzzling phenomena like aha moments", length-scaling'' and entropy dynamics are not disparate occurrences but hallmarks of an emergent reasoning hierarchy.
  arXiv  Detail & Related papers  (2025-09-03T18:52:49Z)
• SEEA-R1: Tree-Structured Reinforcement Fine-Tuning for Self-Evolving Embodied Agents [58.174206358223415]
  Self-Evolving Embodied Agents-R1, or SEEA-R1, is the first reinforcement fine-tuning framework designed for self-evolving embodied agents.<n>We show that SEEA-R1 can support autonomous adaptation and reward-driven self-evolution.
  arXiv  Detail & Related papers  (2025-06-26T18:00:07Z)
• OmniReflect: Discovering Transferable Constitutions for LLM agents via Neuro-Symbolic Reflections [0.8123746895372843]
  We introduce OmniReflect, a reflection-driven framework to improve Large Language Model (LLM) agent performance on complex tasks.<n>We employ Neural, Reflex, and NeuroSymbolic techniques, offering a balance between contextual adaptability and computational efficiency.<n> Empirical results averaged across models show major improvements in task success, with absolute gains of +10.3% on ALFWorld, +23.8% on BabyAI, and +8.3% on PDDL.
  arXiv  Detail & Related papers  (2025-06-20T19:38:21Z)
• LLM Agents for Bargaining with Utility-based Feedback [23.357706450282002]
  We introduce a comprehensive framework centered on utility-based feedback.<n>Our contributions are threefold: (1) BargainArena, a novel benchmark dataset; (2) human-aligned, economically-grounded evaluation metrics inspired by utility theory; and (3) a structured feedback mechanism enabling LLMs to iteratively refine their bargaining strategies.
  arXiv  Detail & Related papers  (2025-05-29T02:07:27Z)
• ReMA: Learning to Meta-think for LLMs with Multi-Agent Reinforcement Learning [53.817538122688944]
  We introduce Reinforced Meta-thinking Agents (ReMA) to elicit meta-thinking behaviors from Reasoning of Large Language Models (LLMs)<n>ReMA decouples the reasoning process into two hierarchical agents: a high-level meta-thinking agent responsible for generating strategic oversight and plans, and a low-level reasoning agent for detailed executions.<n> Empirical results from single-turn experiments demonstrate that ReMA outperforms single-agent RL baselines on complex reasoning tasks.
  arXiv  Detail & Related papers  (2025-03-12T16:05:31Z)
• Retrieval-Augmented Hierarchical in-Context Reinforcement Learning and Hindsight Modular Reflections for Task Planning with LLMs [8.55917897789612]
  We propose Retrieval-Augmented in-context reinforcement Learning (RAHL) for large language models. RAHL decomposes complex tasks into sub-tasks using an LLM-based high-level policy. We show that RAHL can achieve an improvement in performance in 9%, 42%, and 10% in 5 episodes of execution in strong baselines.
  arXiv  Detail & Related papers  (2024-08-12T22:40:01Z)

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.