mlsanityData to 14 Oct 2026

ContextBudget: Budget-Aware Context Management for Long-Horizon Search Agents

arXiv 2604.01664 ↗First posted 2 Apr 2026cs.AIcited by 14

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

Abstract (arXiv)

LLM-based agents show strong potential for long-horizon reasoning, yet their context size is limited by deployment factors (e.g., memory, latency, and cost), yielding a constrained context budget. As interaction histories grow, this induces a trade-off between retaining past information and staying within the context limit. To address this challenge, we propose Budget-Aware Context Management (BACM), which formulates context management as a sequential decision problem with a context budget constraint. It enables agents to assess the available budget before incorporating new observations and decide when and how much of the interaction history to compress. We further develop BACM-RL, an end-to-end curriculum-based reinforcement learning approach that learns compression strategies under varying context budgets. Experiments on compositional multi-objective QA and long-horizon web browsing benchmarks show that BACM-RL consistently outperforms prior methods across model scales and task complexities, achieving over 1.6× gains over strong baselines in high-complexity settings, while maintaining strong advantages as budgets shrink, where most methods exhibit a downward performance trend.

Key facts. Cited by 14 papers in the map (data through 14 Oct 2026). Placed in Agent memory systems, one of 16 directions in LLM agents and tool use. Builds on 28 of its 35 references that are in the map. First posted on arXiv 2 Apr 2026.

Builds on

28 of 35 references.

Closest ancestors · 5

Same direction, newest first

  1. 2025Memory in the Age of AI Agents15 Dec
  2. 2025AgentFold: Long-Horizon Web Agents with Proactive Context Management28 Oct
  3. 2025Memory as Action: Autonomous Context Curation for Long-Horizon Agentic Tasks14 Oct
  4. 2025Scaling Long-Horizon LLM Agent via Context-Folding13 Oct
  5. 2025ReSum: Unlocking Long-Horizon Search Intelligence via Context Summarization16 Sep

Other references · 23

  1. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  2. 2018HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answeringcited by 3,754
  3. 2023Lost in the Middle: How Language Models Use Long Contextscited by 3,455
  4. 2017TriviaQA: A Large Scale Distantly Supervised Challenge Dataset for Reading Comprehensioncited by 2,955
  5. 2024Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameterscited by 1,986
  6. 2025Search-R1: Training LLMs to Reason and Leverage Search Engines with Reinforcement Learningcited by 1,444
  7. 2024RULER: What's the Real Context Size of Your Long-Context Language Models?cited by 1,346
  8. 2020Constructing A Multi-hop QA Dataset for Comprehensive Evaluation of Reasoning Stepscited by 1,282
  9. 2022When Not to Trust Language Models: Investigating Effectiveness of Parametric and Non-Parametric Memoriescited by 1,128
  10. 2022Measuring and Narrowing the Compositionality Gap in Language Modelscited by 1,075
  11. 2021MuSiQue: Multihop Questions via Single-hop Question Compositioncited by 916
  12. 2023MemoryBank: Enhancing Large Language Models with Long-Term Memorycited by 556
11 more
  1. 2025Group-in-Group Policy Optimization for LLM Agent Trainingcited by 497
  2. 2023LongLLMLingua: Accelerating and Enhancing LLMs in Long Context Scenarios via Prompt Compressioncited by 453
  3. 2023Learning to Compress Prompts with Gist Tokenscited by 390
  4. 2023Adapting Language Models to Compress Contextscited by 356
  5. 2023LLMLingua: Compressing Prompts for Accelerated Inference of Large Language Modelscited by 287
  6. 2025MemAgent: Reshaping Long-Context LLM with Multi-Conv RL-based Memory Agentcited by 246
  7. 2025BrowseComp-Plus: A More Fair and Transparent Evaluation Benchmark of Deep-Research Agentcited by 200
  8. 2023Compressing Context to Enhance Inference Efficiency of Large Language Modelscited by 189
  9. 2025MemOS: An Operating System for Memory-Augmented Generation (MAG) in Large Language Modelscited by 73
  10. 2025Steering LLM Thinking with Budget Guidancecited by 33
  11. 2025BudgetThinker: Empowering Budget-aware LLM Reasoning with Control Tokenscited by 29

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 14.

14 papers.

Sort:

By area:

  1. 2026From Question Answering to Task Completion: A Survey on Agent System and Harness Designcited by 12
  2. 2026Self-GC: Self-Governing Context for Long-Horizon LLM Agentscited by 11
  3. 2026Learning When to Remember: Risk-Sensitive Contextual Bandits for Abstention-Aware Memory Retrieval in LLM-Based Coding Agentscited by 3
  4. 2026TokenPilot: Cache-Efficient Context Management for LLM Agentscited by 3
  5. 2026LLM Agents Are Latent Context Managers: Eliciting Self-Managed Context via State Proprioceptioncited by 3
  6. 2026Skim: Speculative Execution for Fast and Efficient Web Agentscited by 1
  7. 2026BAP-SQL: Budget-Aware Observation Planning for Agentic Text-to-SQLcited by 1
  8. 2026Signal-Driven Observation for Long-Horizon Web Agentscited by 0
  9. 2026Context Recycling for Long-Horizon LLM Inferencecited by 0
  10. 2026BudgetBench: A Budget-Tiered Protocol and Pilot Harness for Memory Strategy Evaluation in Local Large Language Model Agentscited by 0
  11. 2026When Can Agents Forget Their Reasoning? ICLR for Long-Horizon Agent Context Compressioncited by 0
  12. 2026ARSM: Auto-Regressive State Machine for Agentic Reasoning Compressioncited by 0
2 more
  1. 2026CADOC: Cache-Aware Dynamic Object Context for Long-Horizon Agentscited by 0
  2. 2026Scaling Multi-Agent Intelligence with Substrate-Centric Coordinationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026ACM: Agentic Context Management for Long Horizon Tasks
  2. 2026ContextPilot: Teaching Agents for Proactive Context Management via Fine-grained RL
  3. 2026FOCUS: Training-Free Decision-Preserving Context Compression for LLM Agents
  4. 2026One Policy, Any Budget: Internalizing Budget-Aware Search via Reinforcement Learning
  5. 2026Escaping the Context Bottleneck: Active Context Curation for LLM Agents via Reinforcement Learning
  6. 2026Learning Agent-Compatible Context Management for Long-Horizon Tasks
  7. 2026Beyond Skill Evolution: Self-Evolving Context Management Policies for Long-Horizon Agent Harnesses
  8. 2026PRO-LONG: Programmatic Memory Enables Long-Horizon Reasoning