Managing Procedural Memory in LLM Agents: Control, Adaptation, and Evaluation

• URL: http://arxiv.org/abs/2606.23127v1
• Date: Mon, 22 Jun 2026 10:14:11 GMT
• Title: Managing Procedural Memory in LLM Agents: Control, Adaptation, and Evaluation
• Abstract summary: AFTER is a benchmark of 382 realistic enterprise tasks spanning six professional roles and 22 procedural skills.<n>Experiments show that procedural memory delivers consistent gains in industrial aggregate performance.<n>These results provide practical guidance for building, evaluating, and deploying procedural memory systems in production agent platforms.
• Score: 6.003738861983597
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Procedural memory is increasingly used to improve LLM agents on recurring workplace tasks, yet its ability to produce reusable skills remains poorly understood. We introduce AFTER, a benchmark of 382 realistic enterprise tasks spanning six professional roles and 22 procedural skills, designed to evaluate how skills transfer across tasks, roles, and model backbones. The benchmark includes controlled evaluation settings for local improvement, cross-task transfer, cross-role transfer, and cross-model generalization. Experiments show that procedural memory delivers consistent gains in industrial workflows: a single refinement round improves aggregate performance by 3.7-6.7 points, while skills evolved from diverse multi-model execution traces achieve 73.1% cross-model test accuracy, outperforming all single-model trace sources. We further find that some skills generalize broadly across tasks and models, whereas others become specialized to role-specific workflows and lose effectiveness under transfer. These results provide practical guidance for building, evaluating, and deploying procedural memory systems in production agent platforms.

Related papers

• Inducing Task Models from Computer-Use Traces [50.6293197430374]
  Task Model Induction (TMI) discovers latent tasks in an unconstrained trace and disentangling concurrent activity.<n>TMI reconstructs 74.9% of the observed execution steps, far more than the strongest workflow induction baseline.<n> skills derived from TMI's task models improve held-out task accuracy by 30.0% over the strongest baseline.
  arXiv  Detail & Related papers  (2026-08-20T17:57:00Z)
• SENTINEL: Failure-Driven Reinforcement Learning for Training Tool-Using Language Model Agents [52.758563996979596]
  Language model agents are increasingly effective in solving realistic tasks through multi-turn tool use.<n>We propose SENTINEL, a failure-driven reinforcement learning framework that turns rollout failures into targeted training tasks.<n>On Tau2-Bench Retail with Qwen3-4B-Thinking-2507, SENTINEL improves Pass1 from 66.4 to 74.9 and outperforms RL on general synthetic tasks across Passk metrics.
  arXiv  Detail & Related papers  (2026-06-11T05:06:50Z)
• SkillEvolBench: Benchmarking the Evolution from Episodic Experience to Procedural Skills [31.23929961213889]
  Large language model (LLM) agents accumulate rich episodic trajectories while solving real-world tasks.<n>We introduce SkillEvolBench, a diagnostic benchmark for evaluating this step from experience reuse to skill formation.<n>We find that current agents often adapt locally but rarely form robust reusable skills.
  arXiv  Detail & Related papers  (2026-05-22T18:23:31Z)
• Claw-Eval-Live: A Live Agent Benchmark for Evolving Real-World Workflows [67.92316850084575]
  We introduce Claw-Eval-Live, a live benchmark for workflow agents.<n>Each release is constructed from public workflow-demand signals.<n>Claw-Eval-Live records execution traces, audit logs, service state, and post-run workspace artifacts.
  arXiv  Detail & Related papers  (2026-04-30T17:23:19Z)
• DARE-bench: Evaluating Modeling and Instruction Fidelity of LLMs in Data Science [31.00353091375463]
  DARE-bench is a benchmark for machine learning modeling and data science instruction.<n>It consists of 6,300 Kaggle-derived tasks and provides both large-scale training data and evaluation sets.<n>Using DARE-bench training tasks for fine-tuning can substantially improve model performance.
  arXiv  Detail & Related papers  (2026-02-27T18:58:57Z)
• Omni-Thinker: Scaling Multi-Task RL in LLMs with Hybrid Reward and Task Scheduling [66.0871543682453]
  We present Omni-Thinker, a unified reinforcement learning framework that scales large language models across diverse tasks.<n>Our scheduler orders tasks according to accuracy backward transfer (BWT), reducing forgetting and improving multi-task performance.
  arXiv  Detail & Related papers  (2025-07-20T01:50:16Z)
• Routine: A Structural Planning Framework for LLM Agent System in Enterprise [10.989149053905587]
  The deployment of agent systems in an enterprise environment is often hindered by several challenges.<n>Common models lack domain-specific process knowledge, leading to disorganized plans, missing key tools, and poor execution stability.<n>This paper introduces Routine, a multi-step agent planning framework designed with a clear structure, explicit instructions, and seamless parameter passing.
  arXiv  Detail & Related papers  (2025-07-19T02:46:19Z)
• Bigger, Regularized, Categorical: High-Capacity Value Functions are Efficient Multi-Task Learners [60.75160178669076]
  We show that the use of high-capacity value models trained via cross-entropy and conditioned on learnable task embeddings addresses the problem of task interference in online reinforcement learning.<n>We test our approach on 7 multi-task benchmarks with over 280 unique tasks, spanning high degree-of-freedom humanoid control and discrete vision-based RL.
  arXiv  Detail & Related papers  (2025-05-29T06:41:45Z)
• Self-Generated In-Context Examples Improve LLM Agents for Sequential Decision-Making Tasks [11.125564622217892]
  Large Language Model agents improve by learning from their own successful experiences without human intervention.<n>Our method constructs and refines a database of self-generated trajectories that serve as in-context examples for future tasks.<n>Our trajectory bootstrapping technique demonstrates that agents can autonomously improve through experience, offering a scalable alternative to labor-intensive knowledge engineering.
  arXiv  Detail & Related papers  (2025-05-01T00:48:12Z)
• Benchmarking Agentic Workflow Generation [80.74757493266057]
  We introduce WorfBench, a unified workflow generation benchmark with multi-faceted scenarios and intricate graph workflow structures.<n>We also present WorfEval, a systemic evaluation protocol utilizing subsequence and subgraph matching algorithms.<n>We observe that the generated can enhance downstream tasks, enabling them to achieve superior performance with less time during inference.
  arXiv  Detail & Related papers  (2024-10-10T12:41:19Z)
• Watch Every Step! LLM Agent Learning via Iterative Step-Level Process Refinement [50.481380478458945]
  Iterative step-level Process Refinement (IPR) framework provides detailed step-by-step guidance to enhance agent training. Our experiments on three complex agent tasks demonstrate that our framework outperforms a variety of strong baselines.
  arXiv  Detail & Related papers  (2024-06-17T03:29:13Z)

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.