# Expert Personas Improve LLM Alignment but Damage Accuracy: Bootstrapping Intent-Based Persona Routing with PRISM arXiv: https://arxiv.org/abs/2603.18507 ยท HTML: https://mlsanity.com/paper/2603.18507/ ## Key facts - Cited by 9 papers in the map (data through 14 Oct 2026). - Placed in Human values and AI feedback alignment, one of 10 directions in LLM alignment / RLHF / preference optimisation. - Builds on 20 of its 31 references that are in the map. - First posted on arXiv 19 Mar 2026. ## Placed in - LLM alignment / RLHF / preference optimisation / [Human values and AI feedback alignment](https://mlsanity.com/alignment/human-values-and-ai-feedback-alignment/) - Synthetic data and self-improvement for LLMs / [LLM synthetic data generation](https://mlsanity.com/synth/llm-synthetic-data-generation/) ## Builds on (most cited first, 20 of 20) - [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023) - [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020) - [Self-Refine: Iterative Refinement with Self-Feedback](https://mlsanity.com/paper/2303.17651/) (2023) - [Self-Instruct: Aligning Language Models with Self-Generated Instructions](https://mlsanity.com/paper/2212.10560/) (2022) - [HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal](https://mlsanity.com/paper/2402.04249/) (2024) - [Chatbot Arena: An Open Platform for Evaluating LLMs by Human Preference](https://mlsanity.com/paper/2403.04132/) (2024) - [A General Language Assistant as a Laboratory for Alignment](https://mlsanity.com/paper/2112.00861/) (2021) - [FrugalGPT: How to Use Large Language Models While Reducing Cost and Improving Performance](https://mlsanity.com/paper/2305.05176/) (2023) - [RouteLLM: Learning to Route LLMs with Preference Data](https://mlsanity.com/paper/2406.18665/) (2024) - [Scaling Synthetic Data Creation with 1,000,000,000 Personas](https://mlsanity.com/paper/2406.20094/) (2024) - [Adapting Language Models to Compress Contexts](https://mlsanity.com/paper/2305.14788/) (2023) - [Beyond Human Data: Scaling Self-Training for Problem-Solving with Language Models](https://mlsanity.com/paper/2312.06585/) (2023) - [LLMLingua-2: Data Distillation for Efficient and Faithful Task-Agnostic Prompt Compression](https://mlsanity.com/paper/2403.12968/) (2024) - [Better Zero-Shot Reasoning with Role-Play Prompting](https://mlsanity.com/paper/2308.07702/) (2023) - [Bias Runs Deep: Implicit Reasoning Biases in Persona-Assigned LLMs](https://mlsanity.com/paper/2311.04892/) (2023) - [In-Context Impersonation Reveals Large Language Models' Strengths and Biases](https://mlsanity.com/paper/2305.14930/) (2023) - [ExpertPrompting: Instructing Large Language Models to be Distinguished Experts](https://mlsanity.com/paper/2305.14688/) (2023) - [From Personas to Talks: Revisiting the Impact of Personas on LLM-Synthesized Emotional Support Conversations](https://mlsanity.com/paper/2502.11451/) (2025) - [PHAnToM: Persona-based Prompting Has An Effect on Theory-of-Mind Reasoning in Large Language Models](https://mlsanity.com/paper/2403.02246/) (2024) - [Persona-Augmented Benchmarking: Evaluating LLMs Across Diverse Writing Styles](https://mlsanity.com/paper/2507.22168/) (2025) ## Built on by (most cited first, 9 of 9) - [When Does Persona Prompting Actually Help? A Retrieval and Metric Analysis of Expert Role Injection in LLMs](https://mlsanity.com/paper/2605.29420/) (2026) - [The Ultimate Tutorial for AI-driven Scale Development in Generative Psychometrics: Releasing AIGENIE from its Bottle](https://mlsanity.com/paper/2603.28643/) (2026) - [Persona Non Grata: Single-Method Safety Evaluation Is Incomplete for Persona-Imbued LLMs](https://mlsanity.com/paper/2604.11120/) (2026) - [Multi-Agent Systems are Mixtures of Experts: Who Becomes an Influencer?](https://mlsanity.com/paper/2605.25929/) (2026) - [The Librarian Who Refused to Code: Model-Dependent Identity Enactment in LLM Code Generation](https://mlsanity.com/paper/2607.17420/) (2026) - [Aligned in Form, Not in Meaning: The Comprehension - Containment Decoupling of LLM Safety in Low-Resource Bangla Derogatory Speech](https://mlsanity.com/paper/2608.02941/) (2026) - [Making Your LLMs More Objective: Stabilizing LLM Safety Behavior Across Traits with Trait-Invariant Safety Tuning](https://mlsanity.com/paper/2608.11705/) (2026) - [mimeo: Compiling Public Expert Corpora into Agent Skills and Testing What Transfers](https://mlsanity.com/paper/2609.00453/) (2026) - [Right Frame, Wrong Rule: Cultural Cues Expose the Financial Knowledge Gap They Were Meant to Close](https://mlsanity.com/paper/2609.00999/) (2026) ## Abstract (arXiv) Persona prompting can steer LLM generation towards a domain-specific tone and pattern. This behavior enables use cases in multi-agent systems where diverse interactions are crucial and human-centered tasks require high-level human alignment. Prior works provide mixed opinions on their utility: some report performance gains when using expert personas for certain domains and their contribution to data diversity in synthetic data creation, while others find near-zero or negative impact on general utility. To fully leverage the benefits of the LLM persona and avoid its harmfulness, a more comprehensive investigation of the mechanism is crucial. In this work, we study how model optimization, task type, prompt length, and placement can impact expert persona effectiveness across instruction-tuned and reasoning LLMs, and provide insight into conditions under which expert personas fail and succeed. Based on our findings, we developed a pipeline to fully leverage the benefits of an expert persona, named PRISM (Persona Routing via Intent-based Self-Modeling), which self-distills an intent-conditioned expert persona into a gated LoRA adapter through a bootstrapping process that requires no external data, models, or knowledge. PRISM enhances human preference and safety alignment on generative tasks while maintaining accuracy on discriminative tasks across all models, with minimal memory and computing overhead.