mlsanityData to 14 Oct 2026

New Model, Old Habits: Behavior Transfer under Model Swapping in Multi-Session LLM Agents

arXiv 2610.19712 ↗First posted 14 Oct 2026cs.LG cs.AIcited by 0

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

Abstract (arXiv)

Multi-session agents can switch between LLMs to balance performance and inference cost. Such switches require transferring a history for continuity, yet their effects on the recipient remain underexplored. Through controlled experiments with multi-session tasks, we uncover behavior transfer: recipients tend to adopt reasoning and action patterns from the models that generated their histories. This effect is most evident with raw-history transfer but is substantially reduced by summarization. Based on these findings, we propose Direction- and Authorship-aware Summarization of History (DASH) that preserves stronger-model behavior while minimizing weak-to-strong transfer via summarization. Experiments with repeated swaps show that DASH outperforms raw-history transfer and naive summarization baselines. These results suggest that effective multi-model orchestration requires considering not only which model to use, but also which behavioral patterns to carry forward.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Agent memory systems, one of 16 directions in LLM agents and tool use. Builds on 26 of its 39 references that are in the map. First posted on arXiv 14 Oct 2026.

Builds on

26 of 39 references.

Closest ancestors · 6

Same direction, newest first

  1. 2025MemEvolve: Meta-Evolution of Agent Memory Systems21 Dec
  2. 2025Memory in the Age of AI Agents15 Dec
  3. 2025How Memory Management Impacts LLM Agents: An Empirical Study of Experience-Following Behavior21 May
  4. 2025LifelongAgentBench: Evaluating LLM Agents as Lifelong Learners17 May
  5. 2025A-MEM: Agentic Memory for LLM Agents17 Feb
  6. 2024A Survey on the Memory Mechanism of Large Language Model based Agents21 Apr

Other references · 20

  1. 2020Language Models are Few-Shot Learnerscited by 32,419
  2. 2023Efficient Memory Management for Large Language Model Serving with PagedAttentioncited by 5,944
  3. 2022Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?cited by 1,434
  4. 2023MemGPT: Towards LLMs as Operating Systemscited by 1,389
  5. 2022WebShop: Towards Scalable Real-World Web Interaction with Grounded Language Agentscited by 1,251
  6. 2021Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order Sensitivitycited by 1,135
  7. 2023FrugalGPT: How to Use Large Language Models While Reducing Cost and Improving Performancecited by 853
  8. 2023Large Language Models Can Be Easily Distracted by Irrelevant Contextcited by 841
  9. 2024RouteLLM: Learning to Route LLMs with Preference Datacited by 682
  10. 2025LLMs Get Lost In Multi-Turn Conversationcited by 429
  11. 2024TravelPlanner: A Benchmark for Real-World Planning with Language Agentscited by 392
  12. 2024GraphRouter: A Graph-based Router for LLM Selectionscited by 165
8 more
  1. 2026DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compressioncited by 76
  2. 2026AgencyBench: Benchmarking the Frontiers of Autonomous Agents in 1M-Token Real-World Contextscited by 28
  3. 2026MTRouter: Cost-Aware Multi-Turn LLM Routing with History-Model Joint Embeddingscited by 10
  4. 2026Inherited Goal Drift: Contextual Pressure Can Undermine Agentic Goalscited by 9
  5. 2026Mitigating Conversational Inertia in Multi-Turn Agentscited by 7
  6. 2026Evaluating Performance Drift from Model Switching in Multi-Turn LLM Systemscited by 6
  7. 2026The Handoff Tax: Continuing Non-Native Trajectories in LLM Agentscited by 6
  8. 2025In-Place Feedback: Reliable Refinement for Multi-Turn Expert-LLM Collaborationcited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026When Does Memory Help Multi-Trajectory Inference for Tool-Use LLM Agents?
  2. 2025What Do LLM Agents Do When Left Alone? Evidence of Spontaneous Meta-Cognitive Patterns
  3. 2026Learn Now, Use Next, Trust Later: Prequential Test-Time Learning for LLM Agents
  4. 2026XferAgent: Cross-Agent Experience Transfer for Self-Improving LLM Agents
  5. 2026Less Context, More Accuracy: A Bi-Temporal Memory Engine for LLM Agents Where a Lean Retrieved Context Beats the Full History
  6. 2026Beyond Compaction: Structured Context Eviction for Long-Horizon Agents
  7. 2026The Epistemics of Agent Memory: Measuring, and Governing, the Consolidation Decision in Long-Horizon LLM Agents
  8. 2026ContextWeaver: Selective and Dependency-Structured Memory Construction for LLM Agents