mlsanityData to 14 Oct 2026

Sleep-time Compute: Beyond Inference Scaling at Test-time

arXiv 2504.13171 ↗First posted 17 Apr 2025cs.AI cs.CLcited by 43

On the map: LLM reasoning → Test-time compute and efficient thinking

Abstract (arXiv)

Scaling test-time compute has emerged as a key ingredient for enabling large language models (LLMs) to solve difficult problems, but comes with high latency and inference cost. We introduce sleep-time compute, which allows models to "think" offline about contexts before queries are presented: by anticipating what queries users might ask and pre-computing useful quantities, we can significantly reduce the compute requirements at test-time. To demonstrate the efficacy of our method, we create modified versions of two reasoning tasks - Stateful GSM-Symbolic and Stateful AIME. We find that sleep-time compute can reduce the amount of test-time compute needed to achieve the same accuracy by ~ 5x on Stateful GSM-Symbolic and Stateful AIME and that by scaling sleep-time compute we can further increase accuracy by up to 13% on Stateful GSM-Symbolic and 18% on Stateful AIME. Furthermore, we introduce Multi-Query GSM-Symbolic, which extends GSM-Symbolic by including multiple related queries per context. By amortizing sleep-time compute across related queries about the same context using Multi-Query GSM-Symbolic, we can decrease the average cost per query by 2.5x. We then conduct additional analysis to understand when sleep-time compute is most effective, finding the predictability of the user query to be well correlated with the efficacy of sleep-time compute. Finally, we conduct a case-study of applying sleep-time compute to a realistic agentic SWE task.

Key facts. Cited by 43 papers in the map (data through 14 Oct 2026). Placed in Test-time compute and efficient thinking, one of 17 directions in LLM reasoning. Builds on 13 of its 18 references that are in the map. First posted on arXiv 17 Apr 2025.

Builds on

13 of 18 references.

Closest ancestors · 2

Same direction, newest first

  1. 2025s1: Simple test-time scaling31 Jan
  2. 2024Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters6 Aug

Other references · 11

  1. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 13,028
  2. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  3. 2023SWE-bench: Can Language Models Resolve Real-World GitHub Issues?cited by 3,373
  4. 2022Fast Inference from Transformers via Speculative Decodingcited by 1,653
  5. 2023MemGPT: Towards LLMs as Operating Systemscited by 1,389
  6. 2024Large Language Monkeys: Scaling Inference Compute with Repeated Samplingcited by 991
  7. 2024Medusa: Simple LLM Inference Acceleration Framework with Multiple Decoding Headscited by 843
  8. 2024GSM-Symbolic: Understanding the Limitations of Mathematical Reasoning in Large Language Modelscited by 550
  9. 2018Blockwise Parallel Decoding for Deep Autoregressive Modelscited by 323
  10. 2024Smaller, Weaker, Yet Better: Training LLM Reasoners via Compute-Optimal Samplingcited by 61
  11. 2022Describing Differences between Text Distributions with Natural Languagecited by 54

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 13; 2026: 30.

43 papers.

Sort:

By area:

  1. 2025Evaluating Memory in LLM Agents via Incremental Multi-Turn Interactionscited by 254
  2. 2025PersonaMem-v2: Towards Personalized Intelligence via Learning Implicit User Personas and Agentic Memorycited by 88
  3. 2025Wait, We Don't Need to "Wait"! Removing Thinking Tokens Improves Reasoning Efficiencycited by 66
  4. 2025Mem-α: Learning Memory Construction via Reinforcement Learningcited by 55
  5. 2026A Survey of Agentic Reasoning for Large Language Models: Towards Recursively Self-Improving and Collective Agentscited by 50
  6. 2025R1-Compress: Long Chain-of-Thought Compression via Chunk Compression and Searchcited by 30
  7. 2025LLM-based Agentic Reasoning Frameworks: A Survey from Methods to Scenarioscited by 26
  8. 2025Get Experience from Practice: LLM Agents with Record & Replaycited by 23
  9. 2026LongMemEval-V2: Evaluating Long-Term Agent Memory Toward Experienced Colleaguescited by 23
  10. 2026Playful Agentic Robot Learningcited by 22
  11. 2025RLAD: Training LLMs to Discover Abstractions for Solving Reasoning Problemscited by 21
  12. 2026Evolving Agents in the Dark: Retrospective Harness Optimization via Self-Preferencecited by 19
31 more
  1. 2025Position: The Real Barrier to LLM Agent Usability is Agentic ROIcited by 12
  2. 2026Auto-Dreamer: Learning Offline Memory Consolidation for Language Agentscited by 10
  3. 2025Revisiting Test-Time Scaling: A Survey and a Diversity-Aware Method for Efficient Reasoningcited by 9
  4. 2026Graph-Native Cognitive Memory for AI Agents: Formal Belief Revision Semantics for Versioned Memory Architecturescited by 9
  5. 2026Autonomous Research Agents: A Survey of AI Scientists and the Verification Gapcited by 9
  6. 2026Neural Chain-of-Thought Search: Searching the Optimal Reasoning Path to Enhance Large Language Modelscited by 8
  7. 2026Position: Modular Memory is the Key to Continual Learning Agentscited by 8
  8. 2026Learning Next Action Predictors from Human-Computer Interactioncited by 7
  9. 2026Panini: Continual Learning in Token Space via Structured Memorycited by 4
  10. 2026IdleSpec: Exploiting Idle Time via Speculative Planning for LLM Agentscited by 4
  11. 2026Anticipate and Learn: Unleashing Idle-Time Compute in Proactive Agentscited by 4
  12. 2026What to Keep, What to Forget: A Rate--Distortion View of Memory Compaction in LLMs and Agentscited by 4
  13. 2026On the Fragility of Self-Improving Agents: Variance, Task Order, and Underspecificationcited by 4
  14. 2025ThinkPilot: Steering Reasoning Models via Automated Think-prefixes Optimizationcited by 3
  15. 2026A Survey on Self-Improving Test-Time Intelligence: Feedback-Driven Adapting, Learning, and Scaling at Inferencecited by 3
  16. 2026Language Models Need Sleep: Learning to Self-Modify and Consolidate Memoriescited by 2
  17. 2026CrystalMem: Elastic Memory for Self-Evolving LLM Agents via Knowledge Crystallizationcited by 2
  18. 2025MIRROR: Converging Cognitive Principles as Computational Mechanisms for AI Reasoningcited by 1
  19. 2025PerCache: Predictive Hierarchical Cache for RAG Applications on Mobile Devicescited by 1
  20. 2026DOVA: Deliberation-First Multi-Agent Orchestration for Autonomous Research Automationcited by 1
  21. 2026Granularity-Regulated Adaptive Computational Efficiency for Optimal Verification in Test-Time Scalingcited by 1
  22. 2026Harness the Memory: A Holistic Evaluation of Memory Substrates in Memory Agentscited by 1
  23. 2026Experience Funnel: A State-Policy Alternating Loop for Self-Evolving Agentscited by 1
  24. 2026BALAR : A Bayesian Agentic Loop for Active Reasoningcited by 0
  25. 2026Do Language Models Need Sleep? Offline Recurrence for Improved Online Inferencecited by 0
  26. 2026The Hitchhiker's Guide to Agentic AI: From Foundations to Systemscited by 0
  27. 2026Second Thought: Reasoning in Parallel as LLM Agents Act and Observecited by 0
  28. 2026DreamBench-SWE: A Multi-Session Memory-Hygiene Benchmark for Software Agentscited by 0
  29. 2026Studying Without a Syllabus: Task-Agnostic Environment Preprocessingcited by 0
  30. 2026Reputation as Community Memory for the Agentic Webcited by 0
  31. 2026Foundations of Proactive Agents: Principles, Technical Layers, and Proactivity-Gymcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025The Energy Cost of Reasoning: Analyzing Energy Usage in LLMs with Test-time Compute
  2. 2025Latency and Token-Aware Test-Time Compute
  3. 2025The Art of Scaling Test-Time Compute for Large Language Models
  4. 2025Are We Scaling the Right Thing? A System Perspective on Test-Time Scaling
  5. 2025Faster and Better LLMs via Latency-Aware Test-Time Scaling
  6. 2025Reasoning on a Budget: A Survey of Adaptive and Controllable Test-Time Compute in LLMs
  7. 2025GSM-Infinite: How Do Your LLMs Behave over Infinitely Increasing Context Length and Reasoning Complexity?
  8. 2025First Finish Search: Efficient Test-Time Scaling in Large Language Models