LikeBench: Evaluating Subjective Likability in LLMs for Personalization

2025-12-15

AI Summary

Current LLM personalization evaluation overemphasizes memory accuracy and information application while neglecting “likability”—a critical subjective dimension. To address this, we propose LikeBench, the first multidimensional, dynamic benchmark for likability assessment. Our method comprises: (i) a novel decomposition of likability into seven diagnosable dimensions—emotional alignment, format matching, knowledge adaptation, etc.; (ii) a fine-grained psychological user simulator modeling preference evolution across multi-turn dialogues; and (iii) a seven-dimensional human annotation protocol coupled with an online preference learning–response adaptation mechanism. Experiments reveal no strong correlation between likability and memory accuracy. DeepSeek-R1 significantly outperforms Qwen3 in likability (+28%), whereas GPT-5 exhibits sharp degradation in robustness to long-horizon noisy dialogues—demonstrating both the necessity and challenge of explicit likability evaluation.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Safety and Robustness
• Humans and AI: Learning Human Values and Preferences

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

A personalized LLM should remember user facts, apply them correctly, and adapt over time to provide responses that the user prefers. Existing LLM personalization benchmarks are largely centered on two axes: accurately recalling user information and accurately applying remembered information in downstream tasks. We argue that a third axis, likability, is both subjective and central to user experience, yet under-measured by current benchmarks. To measure likability holistically, we introduce LikeBench, a multi-session, dynamic evaluation framework that measures likability across multiple dimensions by how much an LLM can adapt over time to a user's preferences to provide more likable responses. In LikeBench, the LLMs engage in conversation with a simulated user and learn preferences only from the ongoing dialogue. As the interaction unfolds, models try to adapt to responses, and after each turn, they are evaluated for likability across seven dimensions by the same simulated user. To the best of our knowledge, we are the first to decompose likability into multiple diagnostic metrics: emotional adaptation, formality matching, knowledge adaptation, reference understanding, conversation length fit, humor fit, and callback, which makes it easier to pinpoint where a model falls short. To make the simulated user more realistic and discriminative, LikeBench uses fine-grained, psychologically grounded descriptive personas rather than the coarse high/low trait rating based personas used in prior work. Our benchmark shows that strong memory performance does not guarantee high likability: DeepSeek R1, with lower memory accuracy (86%, 17 facts/profile), outperformed Qwen3 by 28% on likability score despite Qwen3's higher memory accuracy (93%, 43 facts/profile). Even SOTA models like GPT-5 adapt well in short exchanges but show only limited robustness in longer, noisier interactions.

Problem

• Evaluates subjective likability in LLMs for personalization
• Measures adaptation to user preferences across multiple dimensions
• Decomposes likability into seven diagnostic metrics for analysis

Innovation

• Multi-session dynamic framework for likability evaluation
• Decomposes likability into seven diagnostic metrics
• Uses fine-grained psychologically grounded personas