# The Personalization Mirage: How LLMs Fabricate User Profiles, and Why Self-Monitoring Misleads

arXiv: https://arxiv.org/abs/2608.04570 · HTML: https://mlsanity.com/paper/2608.04570/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 23 of its 38 references that are in the map.
- First posted on arXiv 5 Aug 2026.

## Builds on (most cited first, 23 of 23)

- [DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence](https://mlsanity.com/paper/2606.19348/) (2026)
- [LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory](https://mlsanity.com/paper/2410.10813/) (2024)
- [Know Me, Respond to Me: Benchmarking LLMs for Dynamic User Profiling and Personalized Responses at Scale](https://mlsanity.com/paper/2504.14225/) (2025)
- [How Memory Management Impacts LLM Agents: An Empirical Study of Experience-Following Behavior](https://mlsanity.com/paper/2505.16067/) (2025)
- [PersonaMem-v2: Towards Personalized Intelligence via Learning Implicit User Personas and Agentic Memory](https://mlsanity.com/paper/2512.06688/) (2025)
- [STALE: Can LLM Agents Know When Their Memories Are No Longer Valid?](https://mlsanity.com/paper/2605.06527/) (2026)
- [Measuring What Makes You Unique: Difference-Aware User Modeling for Enhancing LLM Personalization](https://mlsanity.com/paper/2503.02450/) (2025)
- [Mind the Confidence Gap: Overconfidence, Calibration, and Distractor Effects in Large Language Models](https://mlsanity.com/paper/2502.11028/) (2025)
- [Stereotype or Personalization? User Identity Biases Chatbot Recommendations](https://mlsanity.com/paper/2410.05613/) (2024)
- [Actions Speak Louder than Words: Agent Decisions Reveal Implicit Biases in Language Models](https://mlsanity.com/paper/2501.17420/) (2025)
- [HalluDial: A Large-Scale Benchmark for Automatic Dialogue-Level Hallucination Evaluation](https://mlsanity.com/paper/2406.07070/) (2024)
- [Reading Between the Prompts: How Stereotypes Shape LLM's Implicit Personalization](https://mlsanity.com/paper/2505.16467/) (2025)
- [OP-Bench: Benchmarking Over-Personalization for Memory-Augmented Personalized Conversational Agents](https://mlsanity.com/paper/2601.13722/) (2026)
- [Are Large Language Models a Good Replacement of Taxonomies?](https://mlsanity.com/paper/2406.11131/) (2024)
- [Do Language Models Mirror Human Confidence? Exploring Psychological Insights to Address Overconfidence in LLMs](https://mlsanity.com/paper/2506.00582/) (2025)
- [PERMA: Benchmarking Personalized Memory Agents via Event-Driven Preference and Realistic Task Environments](https://mlsanity.com/paper/2603.23231/) (2026)
- [Memoria: A Scalable Agentic Memory Framework for Personalized Conversational AI](https://mlsanity.com/paper/2512.12686/) (2025)
- [AdaMem: Adaptive User-Centric Memory for Long-Horizon Dialogue Agents](https://mlsanity.com/paper/2603.16496/) (2026)
- [LLMs Know They're Wrong and Agree Anyway: The Shared Sycophancy-Lying Circuit](https://mlsanity.com/paper/2604.19117/) (2026)
- [The Personalization Trap: How User Memory Alters Emotional Reasoning in LLMs](https://mlsanity.com/paper/2510.09905/) (2025)
- [Recalling Too Well: Sycophancy Evaluation and Mitigation in Memory-Augmented Models](https://mlsanity.com/paper/2606.10949/) (2026)
- [When Personalization Misleads: Understanding and Mitigating Hallucinations in Personalized LLMs](https://mlsanity.com/paper/2601.11000/) (2026)
- [GRAVITY: Architecture-Agnostic Structured Anchoring for Long-Horizon Conversational Memory](https://mlsanity.com/paper/2605.01688/) (2026)

## Built on by (most cited first, 1 of 1)

- [When to Remember, When to Abstain: Category-Conditioned Retention for Reliable Agent Memory](https://mlsanity.com/paper/2610.07100/) (2026)

## Abstract (arXiv)

Personalized LLMs with persistent memory are increasingly deployed, yet the faithfulness of their user models remains unexamined. We study over-inference (OI): the phenomenon where LLMs fabricate user attributes beyond what evidence supports. We introduce MirageBench, comprising 150 personas balanced across stereotypical, counter-stereotypical, and neutral profiles, 6 personalization tasks spanning an ``imagination gradient'', a four-way faithfulness taxonomy operationalized by an independent judge (validated against a blind human annotator on 400 claims: Cohen's kappa = 0.863 four-class, kappa = 0.900 binary), and a leaderboard of 12 models across 7 families on 143616 judged claims. We find that over-inference is pervasive: every one of the 12 models over-infers 35%--49% of its claims (cross-model mean 41.6%; claim-weighted 41.8%), with no model in this evaluation escaping it. Most strikingly, we surface a Self-Monitoring Inversion: at the model-selection level, models' self-assessed OI is negatively rank-correlated with their judge-measured OI (rho = -0.60, p = 0.044; exploratory, wide bootstrap CI [-0.90, +0.06], n = 12). The models that report the least over-inference tend to be flagged as fabricating the most, so self-reported confidence is a misleading signal for comparing models, even though within a single model self-audit still ranks that model's own claims moderately well (AUROC 0.58--0.83). We further show that OI is task-dependent (27%--59%) and that, in a multi-turn pilot, inferred attributes accumulate approximately linearly with little revision. MirageBench positions external verification, rather than model self-report, as a more reliable foundation for trustworthy personalization.