CORTEXA
← Browse
arxivstat.MLcs.AIcs.LGmath.DG2026-07-02

The Dual Nature of LLM Persona: Aggregated Tendencies and Frame-Dependent Geometry

Yuan Yuan

Evaluations of LLM personas via psychometric questionnaires typically rely on aggregate scores, discarding within-instance correlation structure. We test whether this geometric structure is intrinsic or frame-dependent. Constructing within-instance correlation matrices from IPIP-50 responses, we analyze geometry on SPD manifolds under manipulated question orderings in GPT-4o simulating American and Chinese-American personas. We find that persona expression comprises two dissociable components: aggregated features (Big Five scores) degrade under randomization (21% drop) but are frame-robust; geometric features (SPD manifold) collapse under frame misalignment (42% drop) but recover substantially (to 84%) under shared frames, surpassing aggregated features (76%). This collapse-recovery pattern reveals that persona geometry is not intrinsic but a frame-dependent coordination pattern encoding information invisible to aggregation. Our findings establish a dual-nature framework for LLM personas, frame-dependent geometry versus frame-robust aggregates, necessitating frame-aware evaluation and challenging static trait conceptions.

View free PDFSource page

Related papers

arxivcs.LGcs.AIcs.CLstat.ML2026-06-26

What LLMs explain is not what they believe: Evaluating explanation sufficiency under models' own input beliefs

Nhi Nguyen, Shauli Ravfogel, Rajesh Ranganath

Large language models (LLMs) are increasingly deployed in high-stakes domains, where free-text explanations such as chain-of-thought and post-hoc rationales are used to justify model outputs. Yet it remains unclear whether these explanations are sufficient, i.e., if they contain…

View free PDFSource page
arxivcs.LGcs.AIstat.ML2026-06-29

Behavior Cloning is Not All You Need: The Optimality of On-Policy Distillation for Noisy Expert Feedback

Ved Sriraman, Peihan Liu, Daniel Hsu, Adam Block

Imitation Learning is a natural framework for learning in sequential decision-making systems and has emerged as the dominant paradigm through which we understand language model training. A central puzzle is that, while in theory offline IL can be horizon-free and optimal, in prac…

View free PDFSource page
arxivcs.LGcs.AIcs.ITstat.ML2026-07-19

Kernelized Linear Attention: Breaking the Capacity Wall with Symmetric Cones

Ayoub Ghriss, Sourav Chakraborty

Linear attention promises constant-time recurrent inference but degrades sharply on associative recall. We formulate attention recall as a spherical-packing problem and introduce Kernelized Linear Attention Activations (KATA), a framework whose feature maps are derived from first…

View free PDFSource page
arxivcs.LGcs.AImath.OCstat.ML2026-07-23

A Defense of the Quadratic Model

Alexandru Meterez, Pranav Ajit Nair, Depen Morwani, Cengiz Pehlevan, Sham Kakade, Alex Damian

Due to the complexity of neural network loss landscapes, optimization theory is forced to rely on idealized models, and there is generally a tradeoff between how theoretically tractable the model is, and how accurately it describes the true optimization dynamics. In this work, we…

View free PDFSource page