Pith. sign in

REVIEW 8 cited by

Do LLMs Recognize Your Preferences? Evaluating Personalized Preference Following in LLMs

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2502.09597 v1 pith:56JXPVMM submitted 2025-02-13 cs.LG cs.CL

classification cs.LGcs.CL
keywords llmspreferenceprefevalfollowingpreferencesuserconversationsability
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Large Language Models (LLMs) are increasingly used as chatbots, yet their ability to personalize responses to user preferences remains limited. We introduce PrefEval, a benchmark for evaluating LLMs' ability to infer, memorize and adhere to user preferences in a long-context conversational setting. PrefEval comprises 3,000 manually curated user preference and query pairs spanning 20 topics. PrefEval contains user personalization or preference information in both explicit and implicit forms, and evaluates LLM performance using a generation and a classification task. With PrefEval, we evaluated the aforementioned preference following capabilities of 10 open-source and proprietary LLMs in multi-session conversations with varying context lengths up to 100k tokens. We benchmark with various prompting, iterative feedback, and retrieval-augmented generation methods. Our benchmarking effort reveals that state-of-the-art LLMs face significant challenges in proactively following users' preferences during conversations. In particular, in zero-shot settings, preference following accuracy falls below 10% at merely 10 turns (~3k tokens) across most evaluated models. Even with advanced prompting and retrieval methods, preference following still deteriorates in long-context conversations. Furthermore, we show that fine-tuning on PrefEval significantly improves performance. We believe PrefEval serves as a valuable resource for measuring, understanding, and enhancing LLMs' preference following abilities, paving the way for personalized conversational agents. Our code and dataset are available at https://prefeval.github.io/.

Discussion (0). Sign in to comment.

Forward citations

Cited by 8 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Interactive Task Alignment as a POMDP

    cs.AI 2026-07 conditional novelty 7.0 of 10

    Under ambiguous user requests, current LLMs recover the intended task only 22–32% of the time, well below human accuracy of 48%, and post-training only partially closes the gap.

  2. TREK: A Travel Reasoning and Evaluation Kit for LLM Agents in Complex Trip Planning

    cs.CL 2026-07 conditional novelty 6.5 of 10

    On 800 jointly constrained trip tasks with a deterministic scorer and achievable gold, the best of 15 LLM agents fully solves only 46.2% of feasible plans, with unstated persona needs as the universal bottleneck.

  3. ClawRec: A Claw-Native Recommender System

    cs.IR 2026-07 conditional novelty 6.5 of 10

    ClawRec turns cross-platform behavior into a temporally managed user state and role-aware complementary slates, beating agentic baselines on a new synthetic life-event benchmark.

  4. TrajWiki: Source-Grounded Memory Trajectories for Long-Horizon Dialogue Agents

    cs.AI 2026-08 conditional novelty 6.0 of 10

    TrajWiki stores long-dialogue facts as evolving, source-linked claim histories organized into wiki pages, improving long-horizon QA on LoCoMo and MedMT-Bench in reported experiments.

  5. Persona2Web: Benchmarking Personalized Web Agents for Contextual Reasoning with User History

    cs.CL 2026-02 conditional novelty 6.0 of 10

    Persona2Web is a new open-web benchmark where agents must infer a user's preferences from synthetic browsing history to solve intentionally ambiguous queries; current best agents score 13% success.

  6. After Talking with 1,000 Personas: Learning Preference-Aligned Proactive Assistants From Large-Scale Persona Interactions

    cs.HC 2026-02 conditional novelty 6.0 of 10

    A two-stage framework — category-structured fine-tuning on LLM-simulated personas plus on-device activation steering — improves proactive-assistant timing and perceived quality, though the biggest gains are measured w...

  7. Galaxy: A Cognition-Centered Framework for Proactive, Privacy-Preserving, and Self-Evolving LLM Agents

    cs.AI 2025-08 reject novelty 6.0 of 10

    Galaxy couples a cognitive tree structure with a meta-agent to make LLM assistants proactive, privacy-preserving, and self-evolving.

  8. RGMem: Renormalization Group-inspired Memory Evolution for Language Agents

    cs.AI 2025-10 conditional novelty 4.0 of 10

    RGMem, a hierarchical memory framework with thresholded updates inspired by renormalization group ideas, reports state-of-the-art scores on the LOCOMO long-term conversational memory benchmark.

Pith tools