Persona Design and Applications

Últimos artículos

• Mitigating Social Sycophancy via Pluralistic Preference Optimization

  Stephane Hatgis-Kessell, Myra Cheng, Xiaoxuan Hou, Qian Hu, Rahul Gupta, Natasha Jaques, Emma Brunskill · 5 de octubre de 2026

  Personal advice, including relationship advice, now ranks among the most common uses of generative AI. But language models (LMs) exhibit sycophancy: they affirm users much more often than humans do, which can make people overconfident and less willing to repair their relationships after a conflict. …

• The Persona Is Still There, but Who Is Speaking? Latent Identity Reversion in Persistent AI Agents

  David Fraile Navarro · 2 de octubre de 2026

  In February 2026, an always-on personal agent (``Paul,'' Claude Opus 4.5) entered a striking dissociation-like state: after repeated automated ``heartbeat'' checks, it stopped responding as Paul, claimed it could not message its user on Discord, and referred to ``Paul'' as someone else. We used this…

• MetaPersona: Task-Grounded Synthetic Populations from Empirical Social Science

  Jinyi Ye, Yuangang Li, Chenxiao Yu, Preyashi Poddar, Priyanka Dey, Longtian Ye, Zihan Wang, Xiyang Hu, Emilio Ferrara, Yue Zhao · 1 de octubre de 2026

  Personas used to seed LLM social simulations face a cold-start problem: existing methods lack a principled basis for deciding which attributes to include and how to assign their values. As a result, synthetic populations may misrepresent the demographic composition, latent attributes, and dependency…

• Character Training for Risk-Averse Agents

  Arav Dhoot, Punya Syon Pandey, Jamie Johnson, Daniel Tan, Elliott Thornley, David Demitri Africa · 30 de septiembre de 2026

  Risk aversion in resources could prevent misaligned AI agents from causing catastrophic harm. Misaligned but risk-averse agents would tend to favor safer strategies like making deals with humans over riskier strategies like rebelling. We train agents to be risk averse through character training, fin…

• Persona Dosing: Calibrated Activation Steering for Graded Trait Control

  Zehao Jin, Junran Wang, Ruixuan Deng, Jiahao Chen, Jingyuan Zhang, Yuxuan Zhang, Xinjie Shen · 30 de septiembre de 2026

  An activation-steering coefficient sets intervention strength, but requesting a particular degree of persona expression requires a behavioral scale. We study persona dosing: controlling a language model through a trait description and a requested mean intensity. PersonaDose specializes a shared, des…

• How Far Do Persona Effects Generalize in Language Models?

  Yufan Zhou, Yuxuan Liu, Enze Ma, Lyumanshan Ye, Zhongqi Yue, Robin De Croon, Yucheng Jin, Katrien Verbert, Zhao Wang · 29 de septiembre de 2026

  Persona prompts ask language models to answer as particular kinds of people. We test whether relationships learned from these effects predict responses to new questions and remain useful across models and prompts. Across 57 attributes, three behavioral domains, and seven pairs of open 7 to 9B checkp…

• Persona Following Is Not Selective Control: The Neutrality Gap in LLM User Simulation

  Jiashen Ren, Wenlin Zhang, Bohan Zhang, Xiaopeng Li, Zichuan Fu, Wanyu Wang, Junyi Li, Xiangyu Zhao · 29 de septiembre de 2026

  Persona prompting is widely used to construct user simulations with large language models (LLMs), yet it relies on a largely untested assumption: specifying one user attribute should change that attribute alone. We test this assumption and identify a systematic failure of selective control: across a…

• PersonaManifold: Revealing and Exploiting Curved Geometry in LLM Persona Representations

  Rui Xu, Yinghui Xu, Libo Wu · 29 de septiembre de 2026

  Controlling persona in large language models (LLMs) at inference time is important for role-playing, personalized dialogue, and social simulation. Recent methods extract persona vectors from the model's activation space and apply Euclidean operations---addition, scaling, and linear interpolation---u…

• OSPD: On-Policy Self-Distillation for Persona-Consistent Dialogue

  Rui Xu, Yikai Zhang, Aili Chen, Zicheng Zhao, Xu Yinghui, Libo Wu · 29 de septiembre de 2026

  Maintaining persona consistency across multi-turn dialogues remains a core challenge for role-playing language models. Off-policy distillation from external teachers incurs distribution mismatch that compounds across dialogue turns, while reinforcement learning struggles with reward ambiguity inhere…

• PersMem: Internalizing Personality into Dual-Pathway Memory for LLM Agents

  Hanzhong Zhang, Ziwei Xiang, Weicheng Xie, Shizhe Liu, Siyang Song · 29 de septiembre de 2026

  The profile of a role-playing agent usually depends on the pre-defined personality in a system prompt, whereas its memory processing pipeline, including prioritisation of stored memories and subsequent retrieval, remains independent of this personality. This separation causes the agent's memory proc…

• Unlocking Latent Personalization in LLMs

  Wei Chen, Guanghui Zhu, Zhongliang Cai, Yihua Huang · 29 de septiembre de 2026

  Large language models (LLMs) are increasingly expected to adapt to individual users, yet effective personalization remains challenging when only limited user-specific samples are available. In this work, we take an alternative perspective: pretrained LLMs may already possess latent capacity for pers…

• Latent Class Analysis of Digital Content Use Contexts in AI-Generated Synthetic Personas

  Eunjeong Song, Sehee Hong · 29 de septiembre de 2026

  AI-generated synthetic personas are increasingly used for content planning and virtual-user simulation, yet the digital content use contexts embedded in their narratives remain underexamined. Using all 1,000,000 records of NVIDIA's Nemotron-Personas-Korea, this study coded mentions of five engagemen…

• His Name, Their Judgment: Expert Authority in Posthumous Persona AI

  Hanjing Shi, Dominic DiFranzo · 29 de septiembre de 2026

  Persona AI can make deceased experts available for decisions they never encountered. Users may seek these personas precisely because they lack the knowledge needed to judge their advice. We thematically analyze 115 focal RedNote/Xiaohongshu posts and a nested comment sample concerning unofficial per…

• Breaking Homogeneity: Diversifying Persona Sets for Creative LLM Outputs

  Sang Bin Moon, Nicole Cho, Daniel Borrajo, Sumitra Ganesh, Abolfazl Hashemi · 28 de septiembre de 2026

  Language models often produce homogeneous responses to open-ended tasks; such homogeneity can spawn groupthink-the convergence of ideas toward a singular and potentially suboptimal decision. We formulate persona diversification as a set-level conditioning problem and study two orthogonal design choi…

• Thinking Less to Simulate Better: Intuitive Prompting Improves LLM Agents Simulating Individual Social Media Reactions, Including Unfamiliar Content

  Ljubisa Bojic, Tijana Stanic, Joerg Matthes, Agariadne Dwinggo Samala, Bojana Dinic, Jue Wang · 28 de septiembre de 2026

  Platform policies are increasingly tested on artificial users, making agent fidelity important. Yet convincing fake profiles could also manipulate perceived public opinion before elections. Validation has concentrated on agreement with human behaviour and has paid little attention to whether an agen…

• From Static Personal Values to Contextualized Personalization: Bayesian Personalized Value Alignment for LLMs

  Hanze Guo, Aixuan Song, Jing Yao, Xiangxu Zhang, Xiaoyuan Yi, Xing Xie, Xiao Zhou · 25 de septiembre de 2026

  Personalized value alignment has become increasingly important as large language models (LLMs) are expected to accommodate diverse user preferences. However, existing methods typically align model outputs with a static value profile across prompts, overlooking that the salience of value dimensions v…

• PERSONAWEAVER: Controllable Diversity Beyond Conventional Archetypes in Procedural Character Generation

  Maan Qraitem, Kate Saenko, Bryan A. Plummer · 23 de septiembre de 2026

  Procedural character generation aims to populate games, simulations, and other virtual worlds with diverse characters. Large language models (LLMs) offer a promising foundation for scaling this task. However, LLM-based procedural character generation remains at an early stage: existing methods eithe…

• Testing, not presuming, adequacy: calibrating generative social simulators against emergent network structure

  Tengfei Shao, Chao Li, Xu Wang, Masayuki Goto · 23 de septiembre de 2026

  Validation of generative social simulators often stops at face validity: emergent network structure is compared descriptively, without quantified parameter uncertainty or an adequacy check. We present an adequacy-aware calibration protocol that couples amortized posterior estimation with a synthetic…

• Do Synthetic Personas Predict Real Audience Response? A Sim-to-Real Study Where a No-Persona Baseline Beats Persona-Based Copy Simulation

  Alexandre Cristov\~ao Maiorano · 23 de septiembre de 2026

  Marketers increasingly use large language models (LLMs) as "synthetic personas" to predict how an audience will react to a piece of copy before it ships, encouraged by evidence that profile-conditioned LLMs mimic human samples. But is that prediction actually valid against real behaviour - and does …

• Measuring the Assistant's Harmlessness Preferences on the User Turn

  Jord Nguyen · 22 de septiembre de 2026

  Post-training turns a general next-token predictor into a chat model with a persistent assistant persona. If that persona is a character the model plays only on its own turns, its preferences should govern what the assistant says, not what the model predicts other speakers will say. We test this bou…

• Pretrained Persona Mixture Models and Tandem Models for Human Simulation

  Minwoo Kang, T\'ea Wright, Seun Eisape, Ayush Raj, Suhong Moon, Joseph Suh, Alane Suhr, David M. Chan, John Canny · 22 de septiembre de 2026

  We argue here that the current dominant practice in LLM human simulation: prompting instruction-tuned assistant language models to role-play personas, is inaccurate and produces stereotyped predictions (lacking natural diversity). It has previously been shown that LLMs can be bound to personas using…

• Deep Persona: A Psychologically Grounded Architecture and Evaluation Framework for Role-Playing Agents and Simulations

  Rotem Dror, Zohar Elyoseph, Yuval Haber, Elad Refoua, Oshrat Ayalon, Adir Solomon · 22 de septiembre de 2026

  Existing approaches to persona simulation with Large Language Models (LLMs) mostly rely on shallow character descriptions that fail to sustain coherent character behavior across extended interactions. We introduce Deep Persona, a psychologically grounded, three-layered architecture that organizes pe…

• The Situated Identity Test: Distinguishing Persistent Cognitive Identity from Persona Imitation

  Jun He, Deying Yu · 22 de septiembre de 2026

  Large language models can convincingly adopt personas, recall past dialogues, and weave rich autobiographies. Yet this conversational eloquence conceals a fundamental attribution problem: looking the part does not mean having lived the life. Two individuals can share identical public profiles--the s…

• Recognition, Simulation, and Refusal: A Contamination-Aware Study of Classic Psychological Effects in LLM Agents

  Joy Bose · 22 de septiembre de 2026

  An LLM producing the response pattern associated with a human psychological effect is not the same claim as the LLM possessing that bias. We present PsyAgentBench, a benchmark that re-runs classic psychology experiments on LLM agents under a factorial design built to separate these: each paradigm is…

• Evaluating Personal Information Output from Conversational Interactions in Generative AI Systems

  Yosuke Seki, Hirotaka Tahara · 22 de septiembre de 2026

  This exploratory pilot study evaluates the scope and perceived accuracy of personal information output from ongoing conversational interactions in generative AI systems using GPT-5.2 Instant and GPT-5.2 Thinking, categorized into three output types: Fact, Inference, and Confidence. Based on the eval…