Persona-Conditioned Generation

Follow this topic Ask about this topic

Momentum

12 papers in the last four weeks, up 300% on the four weeks before. 0.1% of all new papers.

Jul 13Week of Sep 28

Latest papers 71

All topics

Decoupling Logic from Persona: Structural Immunity of Edge LLM Agents to Context Pollution
Oct 7, 2026Masaaki Nakatsu, Reno WangLLM Inference LLM Context Management
AO, Inc. / OrbLabs AG · AO, Inc.
Wiki-Talkie: Multilingual Benchmarking of Persona-Based Agents on Real-World Discussions
Oct 6, 2026Dennis Fucci, Andrea Bacciu, Dong Liu +2Multilingual Language Model Evaluation Large Language Model-Based Role-Play Simulation
Amazon
Data-Driven Personas for Survey Simulation: Insights into Simulation Alignment Across Data-Access Regimes
Oct 5, 2026Dongryeol Lee, Weronika Łajewska, Leonardo Perelli +1Computational Social Science User Modeling
Seoul National University · Amazon
Personalized Image Generation with Reasoning and Reflection
Sep 30, 2026Bo Ni, Ngoc N. Tran, Qinwen Ge +10Image Generation Agentic Image Generation
Vanderbilt University · Adobe Systems · University of Maryland, College Park +1
LLM Persona Unlearning
Sep 30, 2026Kemou Li, Zhuan Shi, Qizhou Wang +4Model Editing Persona-Conditioned Generation
State Key Laboratory of Internet of Things for Smart City, University of Macau · Mila – Québec AI Institute · McGill University +3
Calibrated to Whom? Persona and Language Effects on Cultural Values in JEV
Sep 28, 2026Bushra Asseri, Abdulaziz AsseriMultilingual Language Model Evaluation Cross-Cultural Language Model Evaluation
College of Engineering and Advanced Computing, Alfaisal University, Riyadh, Saudi Arabia · Proxa.sa, Riyadh, Saudi Arabia
Persona Dosing: Calibrated Activation Steering for Graded Trait Control
Sep 28, 2026Zehao Jin, Junran Wang, Ruixuan Deng +4Language Model Steering Personality Modeling in Language Models
Georgia Institute of Technology · Zhejiang University · University of British Columbia
Persona Following Is Not Selective Control: The Neutrality Gap in LLM User Simulation
Sep 28, 2026Jiashen Ren, Wenlin Zhang, Bohan Zhang +5Persona-Conditioned Generation User Simulation
City University of Hong Kong · Stanford University
PersonaManifold: Revealing and Exploiting Curved Geometry in LLM Persona Representations
Sep 28, 2026Rui Xu, Yinghui Xu, Libo WuPersonality Modeling in Language Models Persona-Conditioned Generation
Fudan University · Shanghai Innovation Institute
OSPD: On-Policy Self-Distillation for Persona-Consistent Dialogue
Sep 28, 2026Rui Xu, Yikai Zhang, Aili Chen +3Large Language Model-Based Role-Play Simulation Persona Consistency
Fudan University · Shanghai Innovation Institute
Emergi-PersonaOS: A Persona Agent Operating System for Situational Adaptation and Controllable Evolution
Sep 23, 2026Haoluan Fu, Keni Chen, Xinyu Jia +3Persona Consistency Persona-Conditioned Generation
Emergi Lab of Qianzhen Digital Tech, China · Hangzhou Dianzi University, China
PERSONAWEAVER: Controllable Diversity Beyond Conventional Archetypes in Procedural Character Generation
Sep 22, 2026Maan Qraitem, Kate Saenko, Bryan A. PlummerPersonality Modeling in Language Models Procedural Content Generation
Boston University
I Am AdMan: A Pipeline for Automatic Generation of Personalized Advertising Imagery
Sep 14, 2026Victor Kolominsky-Rabas, Leopold Müller, Claudius Budcke +1Conditional Image Generation Image Generation
University of Bayreuth, Universitätsstraße 30, 95447 Bayreuth, Germany · Fraunhofer FIT, Wittelsbacherring 10, 95444 Bayreuth, Germany
A Three-Tier Persona Vector for Controllable User Simulation in Agentic Evaluation
Sep 8, 2026Rahul Khedar, Eshita, Sneha Teja Sree Reddy Thondapu +6LLM Agent Evaluation Multi-Turn Dialogue Evaluation
PayPal AI
Aegix Pulse: A Traceable Three-Stage Architecture for Personalized Content Generation and Context-Preserving Revision
Sep 7, 2026Hongnan Zhao, Shiyu Chen, Zhihao ChenLLM Personalization Evidence-Grounded Generation
1Aegix Insight
Data-Driven Persona-Conditioned Agents for A/B Test Simulation
Sep 1, 2026Ziyad Benomar, Weronika Łajewska, Leonardo Perelli +1Human Behavior Simulation A/b Testing
Amazon · Luxembourg, Luxembourg · Barcelona, Spain
Emotional Labor Strategy Preferences in LLM Personas
Aug 31, 2026Mohammad Saim, Tianyu JiangPersonality Modeling in Language Models Persona-Conditioned Generation
University of Cincinnati
R $^2$ A: Learning Persona Policies Through Persona Representation Learning and Runtime Alignment
Aug 30, 2026Mohan Zhang, Chengsong You, Xiaoyu Cao +4Representation Learning Personality Modeling in Language Models
College of AI, Tsinghua University · 2Independent Researcher
"Many Are My Names": The Anatomy of the Assistant and Its Personas via Sparse Autoencoders
Aug 8, 2026Adelaide Danilov, Aria Nourbakhsh, Oleksandr Marchenko Breneur +1Large Language Model-Based Role-Play Simulation Personality Modeling in Language Models
Department of Computer Science, Faculty of Science, Technology and Medicine, University of Luxembourg
PHASE-Tree: Modeling Character-State Evolution in Long-Horizon Role-Playing Dialogue
Aug 7, 2026Bo Tang, Jianan Yang, Junyi Zhu +7Personality Modeling in Language Models Multi-Turn Dialogue Evaluation
MemTensor (Shanghai) Technology · KU Leuven, Belgium · Zhejiang University +3
Ekova: A Personality-Support Agent for Self-Discovery Dialogue
Aug 1, 2026Yuyan ChenEmotional Support Conversation Persona-Conditioned Generation
ModelsLive Inc.
A Persona-based Rate Action Index
Jul 29, 2026Hayden Helm, Andrew DassoriPersonality Modeling in Language Models Time Series Forecasting
Helivan · Wavelength Capital
Distributional Validity of a Korean Synthetic Persona Panel: Evidence From the Korea Media Panel Survey
Jul 23, 2026Howard Kim, Keuntae ChoPersona-Conditioned Generation Synthetic Data Evaluation
College of AI Convergence, Seoul Cyber University, Seoul 01133, South Korea · Department of Management of Technology, Graduate School, Sungkyunkwan University, Suwon 16419, South Korea · Department of Systems Management Engineering, Sungkyunkwan University, Suwon 16419, South Korea
PrefReward: Learning User Preference Matrix for Personalized Text Generation
Jul 23, 2026Yue Wu, Chengbing Wang, Yimeng Bai +3Personalized Language Models Personalized Text Generation
University of Science and Technology of China · The Chinese University of Hong Kong · National University of Singapore
Evaluating the Effectiveness of Persona Simulation in Opinion Prediction with GPT-4.1
Jul 22, 2026Sarah Y. Li, Ziyu YaoHuman Behavior Simulation Persona-Conditioned Generation
McLean High School · McLean, VA, USA · Department of Computer Science +1
Persona-as-Configuration: Generative Stakeholder Reporting for Agricultural Floods
Jul 20, 2026Oliver Aleksander Larsen, Tiziano Santilli, Francesco Daghero +1Cyber-Physical Systems Structured Output Generation
University of Southern Denmark, Maersk Mc-Kinney Moller Institute, Odense, Denmark
The Librarian Who Refused to Code: Model-Dependent Identity Enactment in LLM Code Generation
Jul 19, 2026Shayell Aharon Salomon, Noam Israel, Ido Safruti +1Personality Modeling in Language Models Code Generation
Bluebear Security
Toward Anthropomorphic Dialogue: A Closed-Loop Framework for Human-Like Chat Generation, Evaluation, and Preference Alignment
Jul 19, 2026Wentao Liu, Siyu Song, Xi Chen +4LLM Alignment Large Language Model-Based Role-Play Simulation
1Shanghai Institute of Innovation · 2East China Normal University · University of Science and Technology of China +3
Distribution-First Population Simulation: Collapse, Calibration, and Recall in Non-WEIRD LLM Persona Modeling
Jul 17, 2026Gurkan OzkanHuman Behavior Simulation Large Language Model-Based Social Simulation
Istanbul Technical University
PACE: Persona Adaptation through Conversational Elicitation in Human-Robot Interaction
Jul 17, 2026Peizhen Li, Longbing Cao, Megani Rajendran +3Human-Robot Interaction Persona-Conditioned Generation
Macquarie University, Australia · NVIDIA, Singapore

Oct 7, 2026, cs.CLJ/K move · Enter open · S save

Decoupling Logic from Persona: Structural Immunity of Edge LLM Agents to Context Pollution

Masaaki Nakatsu, Reno Wang

AO, Inc. / OrbLabs AG · AO, Inc.

LLM Inference LLM Context Management On-Device Language Model Inference Persona-Conditioned Generation

Open paper AskPDF arXiv

Small language-model agents on edge devices must hold a persona and reason correctly at once, inside one context window that fills with conversational history and persona instructions. We study what happens to the logical part of such an agent when that history is long, misleading and persona-heavy (persona-logic interference), and present a Decoupling Architecture (AO-DA) that separates logical inference ("What") from persona expression ("How") into two inference paths on one INT4 base model with hot-swappable LoRA adapters. The logic path receives only the core turn and emits a verifiable structured state (Micro-State); the persona path renders it in character with the full history. In same-base-model ablations on an Apple M2 laptop (Llama-3.1-8B-Instruct and Gemma-3-4B-it, 4-bit; 480 runs over 4 pollution levels x 3 arms x 2 tasks x 2 personas x 5 seeds) we find: (i) the decoupled logic path is structurally invariant to pollution: its prompt stays at 180 (Llama) or 167 (Gemma) tokens while the mixed single-pass prompt grows from 242 to 1,203, and its outputs are byte-identical across levels (40/40); (ii) the mixed single pass degrades monotonically (composite logic score 0.669 to 0.150 on Llama, 0.487 to 0.150 on Gemma), mostly by failing to emit the required structured output (80-95% of runs on Llama, 100% on Gemma at the two highest levels); (iii) with the same pollution fed into the decoupled logic path, the dedicated-adapter, dedicated-format path is still more robust than the single pass on the 8B model (failure 0-20% vs 80-95%; paired

Δ

+0.30 to +0.50, Cliff's

δ

0.50-0.85, Holm-adjusted

p \le 0.03

) but not on the 4B model, where both collapse. Separation costs one extra decode on a topic's first turn (28.2 s vs 18.2 s on Llama) and buys persona hot-swapping in 1.7 ms without re-running the logic path. Code, rubric, fixtures, adapters and logs are released.

Momentum

Related topics

Latest papers 71