Skip to content
AI News HubLIVE
Original source2 min read

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

Summary

PersonaLink distills a user's history into a bounded three-field persona without training, then recursively refines it by scoring a frozen 7B model on a held-out slice of the user's labeled history and keeping only changes that do not regress. With the model fixed, the paper isolates representation from model effects: on LaMP-2 15-way news classification across 200 users, PersonaLink's accuracy (0.745-0.755) is statistically indistinguishable from BM25 retrieval (0.760-0.765). The title indicates that this parity holds for classification but not regression.

SourcearXiv Computational LinguisticsAuthor: JaeHa Yoon, Minjun Park, Seoyeon Kim, Jiwoo Lee, Hyunwoo Choi, Dohyun Kang
Bounded Personas Match Retrieval on Classification but Not Regression for a Frozen Agent
Report an error

The correction channel is not available yet. You can copy the article reference below for later.

Correction instructions
Read article

[Submitted on 25 Jun 2026]

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

View a PDF of the paper titled Bounded Personas Match Retrieval on Classification but Not Regression for a Frozen Agent, by JaeHa Yoon and 5 other authors

View PDF HTML (experimental)

Abstract: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).

Subjects:

Computation and Language (cs.CL)

Cite as: arXiv:2609.02890 [cs.CL]

(or arXiv:2609.02890v1 [cs.CL] for this version)

https://doi.org/10.48550/arXiv.2609.02890

arXiv-issued DOI via DataCite

Submission history

From: Jaeha Yoon [view email] [v1] Thu, 25 Jun 2026 04:02:09 UTC (1,250 KB)

Full-text links:

Access Paper:

View a PDF of the paper titled Bounded Personas Match Retrieval on Classification but Not Regression for a Frozen Agent, by JaeHa Yoon and 5 other authors

View PDF

HTML (experimental)

TeX Source

view license

Current browse context:

cs.CL

new | recent | 2026-09

Change to browse by:

cs

References & Citations

NASA ADS

Google Scholar

Semantic Scholar

Loading...

Data provided by:

Bibliographic Tools

Bibliographic and Citation Tools

Bibliographic Explorer Toggle

Bibliographic Explorer (What is the Explorer?)

Connected Papers Toggle

Connected Papers (What is Connected Papers?)

Litmaps Toggle

Litmaps (What is Litmaps?)

scite.ai Toggle

scite Smart Citations (What are Smart Citations?)

Code, Data, Media

Code, Data and Media Associated with this Article

alphaXiv Toggle

alphaXiv (What is alphaXiv?)

Links to Code Toggle

CatalyzeX Code Finder for Papers (What is CatalyzeX?)

DagsHub Toggle

DagsHub (What is DagsHub?)

GotitPub Toggle

Gotit.pub (What is GotitPub?)

Huggingface Toggle

Hugging Face (What is Huggingface?)

ScienceCast Toggle

ScienceCast (What is ScienceCast?)

Demos

Demos

Replicate Toggle

Replicate (What is Replicate?)

Spaces Toggle

Hugging Face Spaces (What is Spaces?)

Spaces Toggle

TXYZ.AI (What is TXYZ.AI?)

Related Papers

Recommenders and Search Tools

Link to Influence Flower

Influence Flower (What are Influence Flowers?)

Core recommender toggle

CORE Recommender (What is CORE?)

Author

Venue

Institution

Topic

About arXivLabs

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.

Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)

Key points and analysis

Article intelligence

EngineersAdvanced

Key points

  • PersonaLink is training-free and distills a user's history into a bounded, three-field natural-language persona through recursive refinement.
  • Each pass self-evaluates a frozen 7B agent on a held-out slice of the user's own labeled history and keeps rewritten personas only when they do not regress.
  • On LaMP-2 with 200 users and 15-way news categorization, accuracy reaches 0.745-0.755, statistically indistinguishable from BM25 retrieval at 0.760-0.765.
  • The match holds for classification but, as the title states, not for regression with a frozen agent.

Highlights and analysis are generated automatically and may contain errors. Check the original source.