Entity · paper

From Representations to Behaviors: Exploring the Person-Situation-Behavior Triad in LLMs

paperactiveprovisionalfrom-representations-to-behaviors-exploring-the-person-situation-behavior-triad-in-llms-3a8cbd75·1 events·first seen 19h ago

Aliases: From Representations to Behaviors: Exploring the Person-Situation-Behavior Triad in LLMs

Co-occurring entities

Sparse Autoencoder Funder's Personality Triad

More like this (12)

Conversable Complexity: Agentic LLM Collectives as Interpretable Substrates Prompting Complexity: Shortest Prompts for Texts and Behaviors in LLMs From Plausible to Actionable: A Position on LLM Self-Explanations Reason-Mediated Behavioral Models for Auditing LLM Social Simulators Modeling Complex Behaviors: Multi-Personality Composition and Dynamic Switching in Vision-Language Models Attractor States Emerge in Multi-Turn LLM Conversations Revising Context, Shifting Simulated Stance: Auditing LLM-Based Stance Simulation in Online Discussions Understanding the Impact of Linguistic Realization Choices on LLM Stance with Causal Tracing Beyond Third-Person Audits: Situated Interaction Auditing for User-Centered LLM Bias Research From Tokens to States: LLMs as a Special Case of World Models and the Continuous Path Beyond What LLM Agents Say When No One Is Watching: Social Structure and Latent Objective Emergence in Multi-Agent Debates Moral Safety in LLMs: Exposing Performative Compliance with Puzzled Cues

Recent events (1)

6arXiv · cs.CL·19h ago·source ↗

Framework for discovering and controlling personality-like internal representations in LLMs via SAE decomposition

Researchers introduce a framework grounded in Funder's person-situation-behavior triad to identify, intervene on, and validate trait-like internal representations in LLMs. Using sparse autoencoder (SAE) decomposition on contrastive behavior pairs, they locate sparse features corresponding to opposing poles of personality traits, then demonstrate that feature-level interventions produce consistent cross-situational behavioral shifts. Behavioral outcomes on social intelligence tasks show benefit-tradeoff patterns matching human personality research, providing mechanistic evidence that LLMs contain controllable internal structures linking representations to behaviors. The work advances interpretability and controllability of model personality beyond surface-level prompt conditioning.

AI Safety Research Alignment and RLHF Sparse Autoencoder Funder's Personality Triad From Representations to Behaviors: Exploring the Person-Situation-Behavior Triad in LLMs