InComeS is a flexible framework that enhances LLMs'ability to process editing contexts through explicit compression and selection mechanisms, enabling efficient handling of multiple edits without being restricted by the model's context window.
Shuaiyi Li, Zhisong Zhang, Yang Deng et al.· arXiv.org· 1 citation
As large language models are increasingly deployed to simulate diverse human characters, ensuring persona fidelity, defined as the extent to which an agent's behavior consistently reflects the psychological and stylistic characteristics of a target persona, has become a critical requirement. However, existing evaluation paradigms primarily rely on either holistic LLM-based judges, which are prone to"holistic appraisal hallucination'', or static psychometric inventories, which fail to capture the context-dependent fidelity required in dynamic dialogue. To address these limitations, we propose PRISM (Persona Reasoning with Inverse SFL-based Modeling), a psycholinguistically grounded framework that reformulates persona fidelity evaluation as a structured inverse inference task. Inspired by Systemic Functional Linguistics (SFL), PRISM decomposes persona fidelity into three functional dimensions: Task Framing, Interpersonal Stance, and Linguistic Style. It estimates dimension-specific evidence over a persona-conditioned label space and aggregates these signals into an interpretable and auditable evaluation process. Experiments show that PRISM yields more accurate and stable judgements than traditional holistic judging, providing a more reliable framework for persona fidelity evaluation.
Meng-Fan Li, Ze-Sheng Wei, Xuanhua Shi et al.· 1 citation
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
