mlsanityData to 7 Oct 2026

Agentic Context Engineering: Evolving Contexts for Self-Improving Language Models

arXiv 2510.04618 ↗First posted 6 Oct 2025cs.LG cs.AI cs.CLcited by 349

On the map: LLM agents and tool use → Agent memory systems

also in 1 more
Abstract (arXiv)

Large language model (LLM) applications such as agents and domain-specific reasoning increasingly rely on context adaptation: modifying inputs with instructions, strategies, or evidence, rather than weight updates. Prior approaches improve usability but often suffer from brevity bias, which drops domain insights for concise summaries, and from context collapse, where iterative rewriting erodes details over time. We introduce ACE (Agentic Context Engineering), a framework that treats contexts as evolving playbooks that accumulate, refine, and organize strategies through a modular process of generation, reflection, and curation. ACE prevents collapse with structured, incremental updates that preserve detailed knowledge and scale with long-context models. Across agent and domain-specific benchmarks, ACE optimizes contexts both offline (e.g., system prompts) and online (e.g., agent memory), consistently outperforming strong baselines: +10.6% on agents and +8.6% on finance, while significantly reducing adaptation latency and rollout cost. Notably, ACE could adapt effectively without labeled supervision and instead by leveraging natural execution feedback. On the AppWorld leaderboard, ACE matches the top-ranked production-level agent on the overall average and surpasses it on the harder test-challenge split, despite using a smaller open-source model. These results show that comprehensive, evolving contexts enable scalable, efficient, and self-improving LLM systems with low overhead.

Key facts. Cited by 349 papers in the map (data through 7 Oct 2026). Placed in Agent memory systems, one of 16 directions in LLM agents and tool use. Builds on 52 of its 72 references that are in the map. First posted on arXiv 6 Oct 2025.

Builds on

52 of 72 references.

Closest ancestors · 1

Same direction, newest first

  1. 2025A-MEM: Agentic Memory for LLM Agents17 Feb

Other references · 51

  1. 2020Language Models are Few-Shot Learnerscited by 32,199
  2. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,084
  3. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,302
  4. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,160
  5. 2023Judging LLM-as-a-Judge with MT-Bench and Chatbot Arenacited by 8,393
  6. 2022Self-Consistency Improves Chain of Thought Reasoning in Language Modelscited by 5,601
  7. 2023Reflexion: Language Agents with Verbal Reinforcement Learningcited by 4,064
  8. 2021Prefix-Tuning: Optimizing Continuous Prompts for Generationcited by 3,878
  9. 2021The Power of Scale for Parameter-Efficient Prompt Tuningcited by 3,769
  10. 2018HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answeringcited by 3,686
  11. 2023Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflectioncited by 1,800
  12. 2024SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineeringcited by 1,628
38 more
  1. 2020WILDS: A Benchmark of in-the-Wild Distribution Shiftscited by 1,366
  2. 2023Gorilla: Large Language Model Connected with Massive APIscited by 1,302
  3. 2021Improving language models by retrieving from trillions of tokenscited by 1,278
  4. 2019A Comprehensive Survey on Transfer Learningcited by 1,154
  5. 2020In Search of Lost Domain Generalizationcited by 1,042
  6. 2019Machine Unlearningcited by 1,020
  7. 2023Can LLM Already Serve as A Database Interface? A BIg Bench for Large-Scale Database Grounded Text-to-SQLscited by 667
  8. 2024KIVI: A Tuning-Free Asymmetric 2bit Quantization for KV Cachecited by 604
  9. 2023YaRN: Efficient Context Window Extension of Large Language Modelscited by 584
  10. 2022Decomposed Prompting: A Modular Approach for Solving Complex Taskscited by 554
  11. 2025GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learningcited by 449
  12. 2024AppWorld: A Controllable World of Apps and People for Benchmarking Interactive Coding Agentscited by 359
  13. 2024Agent Workflow Memorycited by 342
  14. 2023LegalBench: A Collaboratively Built Benchmark for Measuring Legal Reasoning in Large Language Modelscited by 328
  15. 2024Optimizing Instructions and Demonstrations for Multi-Stage Language Model Programscited by 270
  16. 2023Prompt Cache: Modular Attention Reuse for Low-Latency Inferencecited by 255
  17. 2024TextGrad: Automatic "Differentiation" via Textcited by 250
  18. 2024Rethinking Machine Unlearning for Large Language Modelscited by 247
  19. 2024InfiniGen: Efficient Generative Inference of Large Language Models with Dynamic KV Cache Managementcited by 242
  20. 2024CacheBlend: Fast Large Language Model Serving for RAG with Cached Knowledge Fusioncited by 211
  21. 2024Many-Shot In-Context Learningcited by 203
  22. 2023CacheGen: KV Cache Compression and Streaming for Fast Large Language Model Servingcited by 199
  23. 2025Dynamic Cheatsheet: Test-Time Learning with Adaptive Memorycited by 146
  24. 2022DDXPlus: A New Dataset For Automatic Medical Diagnosiscited by 87
  25. 2022FiNER: Financial Numeric Entity Recognition for XBRL Taggingcited by 47
  26. 2024StreamBench: Towards Benchmarking Continuous Improvement of Language Agentscited by 44
  27. 2025Towards Enterprise-Ready Computer Using Generalist Agentcited by 31
  28. 2025Is Long Context All You Need? Leveraging LLM's Extended Context for NL2SQLcited by 30
  29. 2024Defending Jailbreak Prompts via In-Context Adversarial Gamecited by 27
  30. 2025Agentic Plan Caching: Test-Time Memory for Fast and Cost-Efficient LLM Agentscited by 23
  31. 2025SelfElicit: Your Language Model Secretly Knows Where is the Relevant Evidencecited by 18
  32. 2025The Prompt Alchemist: Automated LLM-Tailored Prompt Optimization for Test Case Generationcited by 16
  33. 2025Adaptive Self-improvement LLM Agentic System for ML Library Developmentcited by 16
  34. 2017Overcoming data scarcity with transfer learningcited by 14
  35. 2023Generating Data for Symbolic Language with Large Language Modelscited by 12
  36. 2025FinLoRA: Benchmarking LoRA Methods for Fine-Tuning LLMs on Financial Datasetscited by 11
  37. 2025Putting It All into Context: Simplifying Agents with LCLMscited by 9
  38. 2024LIFT: Improving Long Context Understanding Through Long Input Fine-Tuningcited by 6

50 most cited of 51 All 51 (JSON)

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 26; 2026: 323.

349 papers.

Sort:

By area:

  1. 2025Memory in the Age of AI Agentscited by 285
  2. 2026Meta-Harness: End-to-End Optimization of Model Harnessescited by 248
  3. 2026Agentic Harness Engineering: Observability-Driven Automatic Evolution of Coding-Agent Harnessescited by 119
  4. 2026Self-Harness: Harnesses That Improve Themselvescited by 79
  5. 2026SkillX: Automatically Constructing Skill Knowledge Bases for Agentscited by 69
  6. 2025ThetaEvolve: Test-time Learning on Open Problemscited by 52
  7. 2025FLEX: Continuous Agent Evolution via Forward Learning from Experiencecited by 51
  8. 2026Agent World Model: Infinity Synthetic Environments for Agentic Reinforcement Learningcited by 51
  9. 2026A Survey of Agentic Reasoning for Large Language Models: Towards Recursively Self-Improving and Collective Agentscited by 47
  10. 2026AdaEvolve: Adaptive LLM Driven Zeroth-Order Optimizationcited by 46
  11. 2026EvoX: Meta-Evolution for Automated Discoverycited by 45
  12. 2025Continuum: Efficient and Robust Multi-Turn LLM Agent Scheduling with KV Cache Time-to-Livecited by 43
38 more
  1. 2026Meta Context Engineering via Agentic Skill Evolutioncited by 42
  2. 2026Vision-as-Inverse-Graphics Agent via Interleaved Multimodal Reasoningcited by 40
  3. 2026A Survey of Agent Memory in the Second Half: Towards Self-Evolving and Long-Horizon Agentscited by 40
  4. 2026Code as Agent Harnesscited by 40
  5. 2026Rethinking the Evaluation of Harness Evolution for Agentscited by 39
  6. 2025Measuring Agents in Productioncited by 35
  7. 2025CASCADE: Cumulative Agentic Skill Creation through Autonomous Development and Evolutioncited by 33
  8. 2026Dive into Claude Code: The Design Space of Today's and Future AI Agent Systemscited by 33
  9. 2026Harness Updating Is Not Harness Benefit: Disentangling Evolution Capabilities in Self-Evolving LLM Agentscited by 28
  10. 2026Toward Ultra-Long-Horizon Agentic Science: Cognitive Accumulation for Machine Learning Engineeringcited by 26
  11. 2026Stop Comparing LLM Agents Without Disclosing the Harnesscited by 25
  12. 2025AccelOpt: A Self-Improving LLM Agentic System for AI Accelerator Kernel Optimizationcited by 21
  13. 2026From Failed Trajectories to Reliable LLM Agents: Diagnosing and Repairing Harness Flawscited by 21
  14. 2026RealMem: Benchmarking LLMs in Real-World Memory-Driven Interactioncited by 20
  15. 2025Adaptation of Agentic AI: A Survey of Post-Training, Memory, and Skillscited by 19
  16. 2026Evaluating AGENTS.md: Are Repository-Level Context Files Helpful for Coding Agents?cited by 19
  17. 2026From Context to Skills: Can Language Models Learn from Context Skillfully?cited by 19
  18. 2025Resolving Evidence Sparsity: Agentic Context Engineering for Long-Document Understandingcited by 18
  19. 2026SkillsVote: Lifecycle Governance of Agent Skills from Collection, Recommendation to Evolutioncited by 18
  20. 2026A Comprehensive Survey on Agent Skills: Taxonomy, Techniques, and Applicationscited by 17
  21. 2026Self-Improvements in Modern Agentic Systems: A Surveycited by 17
  22. 2025SCOPE: Prompt Evolution for Enhancing Agent Effectivenesscited by 16
  23. 2026Trajectory-Informed Memory Generation for Self-Improving Agent Systemscited by 16
  24. 2026Co-Harness: Co-Evolving Harnesses and Model Weights for LLM Agentscited by 16
  25. 2025DocLens : A Tool-Augmented Multi-Agent Framework for Long Visual Document Understandingcited by 15
  26. 2026EvoMemBench: Benchmarking Agent Memory from a Self-Evolving Perspectivecited by 14
  27. 2026Position: Agentic Evolution is the Path to Evolving LLMscited by 13
  28. 2026TrajAD: Trajectory Anomaly Detection for Trustworthy LLM Agentscited by 13
  29. 2026SimWorld Studio: Automatic Environment Generation with Evolving Coding Agent for Embodied Agent Learningcited by 13
  30. 2025AI Meets Brain: Memory Systems from Cognitive Neuroscience to Autonomous Agentscited by 12
  31. 2026Evaluating Memory Structure in LLM Agentscited by 12
  32. 2026On Safety Risks in Experience-Driven Self-Evolving Agentscited by 12
  33. 2026SkillEvolver: Skill Learning as a Meta-Skillcited by 12
  34. 2026Evo-Harness: Context-to-Harness Skill Compilation for Self-Evolving Agentscited by 12
  35. 2025Agentic Learner with Grow-and-Refine Multimodal Semantic Memorycited by 11
  36. 2026Toward Efficient Agents: Memory, Tool learning, and Planningcited by 11
  37. 2026Continual Learning Bench: Evaluating Frontier AI Systems in Real-World Stateful Environmentscited by 11
  38. 2026SkillAudit: Ground-Truth-Free Skill Evolution via Paired Trajectory Auditingcited by 11

50 most cited of 349 All 349 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2026ACE: Pluggable Adaptive Context Elasticizer across Agents
  2. 2025Towards more Contextual Agents: An extractor-Generator Optimization Framework
  3. 2025A Survey of Context Engineering for Large Language Models
  4. 2026ContextPilot: Teaching Agents for Proactive Context Management via Fine-grained RL
  5. 2025PAACE: A Plan-Aware Automated Agent Context Engineering Framework
  6. 2025ACON: Optimizing Context Compression for Long-horizon LLM Agents
  7. 2026ACM: Agentic Context Management for Long Horizon Tasks
  8. 2026ContextPipe: Database-Inspired Context Assembly for Long-Horizon Agents