Imagen de Portada

Tracing Persona Vectors Through LLM Pretraining

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Moskvoretskii, Viktor, Glandorf, Dominik, Moreira, Jorge Medina, Käser, Tanja, West, Robert
Formato:	Preprint
Publicado:	2026
Materias:	Computation and Language Artificial Intelligence
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

Scalable and Explainable Learner-Video Interaction Prediction using Multimodal Large Language Models
por: Glandorf, Dominik, et al.
Publicado: (2026)

Where is the Mind? Persona Vectors and LLM Individuation
por: Beckmann, Pierre, et al.
Publicado: (2026)

SCRIBE: Structured Chain Reasoning for Interactive Behaviour Explanations using Tool Calling
por: Fawzi, Fares, et al.
Publicado: (2025)

Evolutionary Search for Automated Design of Uncertainty Quantification Methods
por: Seleznyov, Mikhail, et al.
Publicado: (2026)

Leveraging LLM Parametric Knowledge for Fact Checking without Retrieval
por: Vazhentsev, Artem, et al.
Publicado: (2026)

Next Token Knowledge Tracing: Exploiting Pretrained LLM Representations to Decode Student Behaviour
por: Norris, Max, et al.
Publicado: (2025)

Moving Beyond Review: Applying Language Models to Planning and Translation in Reflection
por: Neshaei, Seyed Parsa, et al.
Publicado: (2026)

ClickSight: Interpreting Student Clickstreams to Reveal Insights on Learning Strategies via LLMs
por: Radmehr, Bahar, et al.
Publicado: (2025)

The Impact of Steering Large Language Models with Persona Vectors in Educational Applications
por: Wu, Yongchao, et al.
Publicado: (2026)

Narrow Finetuning Leaves Clearly Readable Traces in Activation Differences
por: Minder, Julian, et al.
Publicado: (2025)

BILLY: Steering Large Language Models via Merging Persona Vectors for Creative Generation
por: Pai, Tsung-Min, et al.
Publicado: (2025)

Polypersona: Persona-Grounded LLM for Synthetic Survey Responses
por: Dash, Tejaswani, et al.
Publicado: (2025)

Bullying the Machine: How Personas Increase LLM Vulnerability
por: Xu, Ziwei, et al.
Publicado: (2025)

daVinci-LLM:Towards the Science of Pretraining
por: Qin, Yiwei, et al.
Publicado: (2026)

Beyond Discrete Personas: Personality Modeling Through Journal Intensive Conversations
por: Pal, Sayantan, et al.
Publicado: (2024)

A Systematic Analysis of the Impact of Persona Steering on LLM Capabilities
por: Chen, Jiaqi, et al.
Publicado: (2026)

Algorithmic Fragility and Persona Bias in LLM-Generated Autistic Communication
por: Rizvi, Naba, et al.
Publicado: (2026)

Sydney Telling Fables on AI and Humans: A Corpus Tracing Memetic Transfer of Persona between LLMs
por: Milička, Jiří, et al.
Publicado: (2026)

Crosscoding Through Time: Tracking Emergence & Consolidation Of Linguistic Representations Throughout LLM Pretraining
por: Bayazit, Deniz, et al.
Publicado: (2025)

Optimizing Pretraining Data Mixtures with LLM-Estimated Utility
por: Held, William, et al.
Publicado: (2025)

Simulating Misinformation Vulnerabilities With Agent Personas
por: Farr, David, et al.
Publicado: (2025)

Through the Judge's Eyes: Inferred Thinking Traces Improve Reliability of LLM Raters
por: Zhang, Xingjian, et al.
Publicado: (2025)

Pay What LLM Wants: Can LLM Simulate Economics Experiment with 522 Real-human Persona?
por: Choi, Junhyuk, et al.
Publicado: (2025)

TalkDep: Clinically Grounded LLM Personas for Conversation-Centric Depression Screening
por: Wang, Xi, et al.
Publicado: (2025)

German General Social Survey Personas: A Survey-Derived Persona Prompt Collection for Population-Aligned LLM Studies
por: Rupprecht, Jens, et al.
Publicado: (2025)

Exploring Persona-dependent LLM Alignment for the Moral Machine Experiment
por: Kim, Jiseon, et al.
Publicado: (2025)

Revisiting the UID Hypothesis in LLM Reasoning Traces
por: Gwak, Minju, et al.
Publicado: (2025)

A Continued Pretrained LLM Approach for Automatic Medical Note Generation
por: Yuan, Dong, et al.
Publicado: (2024)

SensorPersona: An LLM-Empowered System for Continual Persona Extraction from Longitudinal Mobile Sensor Streams
por: Yang, Bufang, et al.
Publicado: (2026)

Not All Personas Are Worth It: Culture-Reflective Persona Data Augmentation
por: Han, Ji-Eun, et al.
Publicado: (2025)

Beyond Cooperative Simulators: Generating Realistic User Personas for Robust Evaluation of LLM Agents
por: Chopra, Harshita, et al.
Publicado: (2026)

Tracing the Representation Geometry of Language Models from Pretraining to Post-training
por: Li, Melody Zixuan, et al.
Publicado: (2025)

Temperature and Persona Shape LLM Agent Consensus With Minimal Accuracy Gains in Qualitative Coding
por: Borchers, Conrad, et al.
Publicado: (2025)

The Persona Paradox: Medical Personas as Behavioral Priors in Clinical Language Models
por: Abdullahi, Tassallah, et al.
Publicado: (2026)

PersonaMatrix: A Recipe for Persona-Aware Evaluation of Legal Summarization
por: Pang, Tsz Fung, et al.
Publicado: (2025)

Chinese Tiny LLM: Pretraining a Chinese-Centric Large Language Model
por: Du, Xinrun, et al.
Publicado: (2024)

ReasonOps: Operator Segmentation for LLM Reasoning Traces
por: Lee, Daniel, et al.
Publicado: (2026)

$\texttt{BluePrint}$: A Social Media User Dataset for LLM Persona Evaluation and Training
por: Bück-Kaeffer, Aurélien, et al.
Publicado: (2025)

Generalization v.s. Memorization: Tracing Language Models' Capabilities Back to Pretraining Data
por: Wang, Xinyi, et al.
Publicado: (2024)

Output Embedding Centering for Stable LLM Pretraining
por: Stollenwerk, Felix, et al.
Publicado: (2026)