mlsanityData to 14 Oct 2026

Bounded Personas Match Retrieval on Classification but Not Regression for a Frozen Agent

arXiv 2609.02890 ↗First posted 25 Jun 2026cs.CLcited by 0

Not on the map. Why?

Abstract (arXiv)

A personalized language agent must convert a user's interaction history into behavior on each new request at inference time. Two strategies dominate. Retrieval pulls a few of the user's most relevant past items into the prompt, which is accurate but pays a per-query selection and context cost that grows with the history. Distillation instead compresses the history once into a compact natural-language persona, which is bounded, query-independent, and interpretable, but is widely assumed to sacrifice accuracy. Whether, and on which tasks, a distilled persona can match retrieval has not been characterized cleanly. We introduce PersonaLink, a training-free method that distills a user's history into a bounded three-field persona and recursively refines it: each pass self-evaluates the frozen agent on a held-out slice of the user's own labeled history, rewrites the persona from its errors, and keeps the result only when it does not regress on that slice. Because every comparison shares one frozen 7B backbone and differs only in what is placed in context, the design isolates the effect of representation from that of the model. The result is a clear task-type asymmetry. On 200 users of LaMP-2 (15-way news categorization), PersonaLink reaches 0.745-0.755 accuracy, statistically indistinguishable from BM25 retrieval (0.760-0.765).

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 80 of its 83 references that are in the map. First posted on arXiv 25 Jun 2026.

Builds on

80 of 83 references.

References · 80

  1. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,297
  2. 2019Sentence-BERT: Sentence Embeddings using Siamese BERT-Networkscited by 9,170
  3. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,695
  4. 2022Self-Consistency Improves Chain of Thought Reasoning in Language Modelscited by 5,708
  5. 2023Reflexion: Language Agents with Verbal Reinforcement Learningcited by 4,276
  6. 2020Dense Passage Retrieval for Open-Domain Question Answeringcited by 3,824
  7. 2023Generative Agents: Interactive Simulacra of Human Behaviorcited by 3,568
  8. 2023Self-Refine: Iterative Refinement with Self-Feedbackcited by 3,442
  9. 2023Tree of Thoughts: Deliberate Problem Solving with Large Language Modelscited by 3,211
  10. 2022Constitutional AI: Harmlessness from AI Feedbackcited by 3,050
  11. 2020REALM: Retrieval-Augmented Language Model Pre-Trainingcited by 2,141
  12. 2024Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameterscited by 1,986
38 more
  1. 2023Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflectioncited by 1,839
  2. 2023MemGPT: Towards LLMs as Operating Systemscited by 1,389
  3. 2025s1: Simple test-time scalingcited by 1,341
  4. 2021Improving language models by retrieving from trillions of tokenscited by 1,293
  5. 2021Unsupervised Dense Information Retrieval with Contrastive Learningcited by 1,216
  6. 2020Leveraging Passage Retrieval with Generative Models for Open Domain Question Answeringcited by 1,210
  7. 2021What Makes Good In-Context Examples for GPT-3?cited by 1,195
  8. 2020ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERTcited by 1,178
  9. 2022Large Language Models Are Human-Level Prompt Engineerscited by 1,045
  10. 2023Large Language Models Cannot Self-Correct Reasoning Yetcited by 1,017
  11. 2022STaR: Bootstrapping Reasoning With Reasoningcited by 940
  12. 2023ExpeL: LLM Agents Are Experiential Learnerscited by 876
  13. 2023Large Language Models as Optimizerscited by 864
  14. 2020Approximate Nearest Neighbor Negative Contrastive Learning for Dense Text Retrievalcited by 849
  15. 2023DSPy: Compiling Declarative Language Model Calls into Self-Improving Pipelinescited by 832
  16. 2018Personalizing Dialogue Agents: I have a dog, do you have pets too?cited by 828
  17. 2023REPLUG: Retrieval-Augmented Black-Box Language Modelscited by 792
  18. 2023CRITIC: Large Language Models Can Self-Correct with Tool-Interactive Critiquingcited by 783
  19. 2023In-Context Retrieval-Augmented Language Modelscited by 702
  20. 2024Self-Rewarding Language Modelscited by 681
  21. 2022Large Language Models Can Self-Improvecited by 664
  22. 2021Learning To Retrieve Prompts for In-Context Learningcited by 648
  23. 2023MemoryBank: Enhancing Large Language Models with Long-Term Memorycited by 556
  24. 2023Promptbreeder: Self-Referential Self-Improvement Via Prompt Evolutioncited by 551
  25. 2025GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learningcited by 542
  26. 2023LaMP: When Large Language Models Meet Personalizationcited by 437
  27. 2025Agentic Context Engineering: Evolving Contexts for Self-Improving Language Modelscited by 398
  28. 2024Agent Workflow Memorycited by 383
  29. 2023A Survey on Large Language Models for Recommendationcited by 379
  30. 2023Character-LLM: A Trainable Agent for Role-Playingcited by 339
  31. 2022Demonstrate-Search-Predict: Composing retrieval and language models for knowledge-intensive NLPcited by 318
  32. 2021SPLADE: Sparse Lexical and Expansion Model for First Stage Rankingcited by 286
  33. 2023Personalized Soups: Personalized Large Language Model Alignment via Post-hoc Parameter Mergingcited by 280
  34. 2024When Can LLMs Actually Correct Their Own Mistakes? A Critical Survey of Self-Correction of LLMscited by 240
  35. 2024From Persona to Personalization: A Survey on Role-Playing Language Agentscited by 212
  36. 2024Two Tales of Persona in LLMs: A Survey of Role-Playing and Personalizationcited by 208
  37. 2024Meta-Rewarding Language Models: Self-Improving Alignment with LLM-as-a-Meta-Judgecited by 204
  38. 2019TransferTransfo: A Transfer Learning Approach for Neural Network Based Conversational Agentscited by 188

50 most cited of 80 All 80 (JSON)

Built on by

Not cited yet.