Learning What to Remember and What to Internalize in LLM Self-Evolution via Adaptive Memory-Parameter Coordination

• URL: http://arxiv.org/abs/2608.01234v1
• Date: Sun, 02 Aug 2026 13:39:01 GMT
• Title: Learning What to Remember and What to Internalize in LLM Self-Evolution via Adaptive Memory-Parameter Coordination
• Abstract summary: We present COVE, a unified agent self-evolution framework.<n>It combines harness-based and parameter-based learning through task-aware routing, stage-aware scheduling, and knowledge optimization.<n>COVE outperforms single-channel evolution strategies, demonstrating more robust and efficient improvement under changing environments.
• Score: 40.98046592234726
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large language model agents increasingly operate in dynamic environments where tool interfaces, APIs, and user requirements change after deployment. Existing self-evolution methods mainly follow two paradigms: harness-based approaches, which externalize feedback into editable memories or skills for rapid adaptation, and parameter-based approaches, which internalize experience into model parameters for deeper capability improvement. However, using either mechanism alone creates a trade-off between flexibility and performance. This paper asks how an agent can coordinate both channels to achieve robust self-evolution. We present COVE, a unified agent self-evolution framework that combines harness-based and parameter-based learning through task-aware routing, stage-aware scheduling, and knowledge optimization. Through this design, COVE treats self-evolution not as indiscriminate accumulation of experience, but as a coordinated process that matches tasks and knowledge types to appropriate learning mechanisms. Experiments across multiple task categories show that COVE outperforms single-channel evolution strategies, demonstrating more robust and efficient improvement under changing environments.

Related papers

• Self-Evolving Memory for Generative Recommendation [100.15207958893029]
  Generative recommendation has emerged as a promising end-to-end paradigm for personalized recommendation.<n>Existing evolving strategies, such as continual retraining and distillation-based adaptation, directly update the shared model parameters.<n>We propose a self-evolving memory paradigm for generative recommendation, aiming to enable effective evolution across heterogeneous behavioral patterns.
  arXiv  Detail & Related papers  (2026-09-14T14:00:38Z)
• Self-Improving Large Language Models via Progressive Experience Evolution [52.03479747358687]
  We propose textbfSPEE (textbfSelf-textbfProgressive textbfExperience textbfEvolution), a unified post-training framework that sequentially performs explicit experience evolution followed by implicit policy optimization.<n> Experiments on five mathematical reasoning benchmarks demonstrate that SPEE consistently outperforms both test-time and training-time self-evolution baselines across three model scales.
  arXiv  Detail & Related papers  (2026-08-03T12:27:32Z)
• Differentiable Mixture-of-Agents Incentivizes Swarm Intelligence of Large Language Models [17.7409616106227]
  Differentiable Mixture-of-Agents (DMoA) is a self-evolving multi-agent framework that enables elastic and adaptive agent collaboration during inference.<n>We show that DMoA achieves state-of-the-art performance while exhibiting strong efficiency, robustness, and ensembling capabilities.
  arXiv  Detail & Related papers  (2026-05-15T07:54:46Z)
• Heterogeneous Multi-Agent Reinforcement Learning with Attention for Cooperative and Scalable Feature Transformation [21.732611237889326]
  Feature transformation enhances downstream task performance by generating informative features through mathematical feature crossing.<n>Recent works employ reinforcement learning to enhance traditional approaches through a more effective trial-and-error way.<n>We propose a novel heterogeneous multi-agent RL framework to enable cooperative and scalable feature transformation.
  arXiv  Detail & Related papers  (2025-11-26T21:45:38Z)
• Fints: Efficient Inference-Time Personalization for LLMs with Fine-Grained Instance-Tailored Steering [49.212940215720884]
  We propose a steering framework that generates sample-level interference from user data and injects it into the model's forward pass for personalized adaptation.<n>Our method significantly enhances personalization performance in fast-shifting environments while maintaining robustness across varying interaction modes and context lengths.
  arXiv  Detail & Related papers  (2025-10-31T06:01:04Z)
• A Survey of Self-Evolving Agents: On Path to Artificial Super Intelligence [87.08051686357206]
  Large Language Models (LLMs) have demonstrated strong capabilities but remain fundamentally static.<n>As LLMs are increasingly deployed in open-ended, interactive environments, this static nature has become a critical bottleneck.<n>This survey provides the first systematic and comprehensive review of self-evolving agents.
  arXiv  Detail & Related papers  (2025-07-28T17:59:05Z)
• Automation and Feature Selection Enhancement with Reinforcement Learning (RL) [0.0]
  Reinforcement learning integrated with decision tree improves feature knowledge, state representation and selection efficiency.<n>Monte Carlo-based reinforced feature selection(MCRFS), a single-agent feature selection method reduces computational burden.<n>A dual-agent RL framework is also introduced that collectively selects features and instances, capturing the interactions between them.
  arXiv  Detail & Related papers  (2025-03-15T04:30:55Z)
• Investigate-Consolidate-Exploit: A General Strategy for Inter-Task Agent Self-Evolution [92.84441068115517]
  Investigate-Consolidate-Exploit (ICE) is a novel strategy for enhancing the adaptability and flexibility of AI agents. ICE promotes the transfer of knowledge between tasks for genuine self-evolution. Our experiments on the XAgent framework demonstrate ICE's effectiveness, reducing API calls by as much as 80%.
  arXiv  Detail & Related papers  (2024-01-25T07:47:49Z)
• Meta-Reinforcement Learning for Adaptive Control of Second Order Systems [3.131740922192114]
  In process control, many systems have similar and well-understood dynamics, which suggests it is feasible to create a generalizable controller through meta-learning. We formulate a meta reinforcement learning (meta-RL) control strategy that takes advantage of known, offline information for training, such as a model structure. A key design element is the ability to leverage model-based information offline during training, while maintaining a model-free policy structure for interacting with new environments.
  arXiv  Detail & Related papers  (2022-09-19T18:51:33Z)

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.