Institution profile

Dialpad Inc.

Industry researchnorthamerica · us
Official website
Research library18linked papers
Opportunities0open roles
Selected work

Representative Papers

Back in Style: A Sociolinguistic Approach to Authoring and Measuring Persona Fidelity in User Simulation

Oct 07, 2026

This study addresses the low fidelity of user simulators and their reliance on costly, subjective LLM-based evaluation by proposing an objective assessment framework grounded in sociolinguistics. Methodologically, personality is reconceptualized as observable linguistic style rather than a label prediction task, enabling deterministic measurement through specific style rates. By integrating stylometric and lexicon-based analysis techniques, the framework introduces model-free validation tools for objective diagnostics. Experimental results demonstrate that this approach significantly enhances both style adherence and discriminability across most models, offering a novel pathway for precisely identifying simulation errors and generating diverse user profiles.

0 citationsRead paper

Prompts versus Rules: Auditing and Controlling Speech Naturalness Behaviors in Voice User Simulators

Oct 07, 2026

This study addresses the disconnect between configured naturalness behaviors and actual outputs in spoken user simulators, which compromises evaluation validity. To this end, we propose a linguistics-based deterministic rule injection algorithm alongside a naturalness behavior auditing framework. By systematically comparing large language model (LLM) prompt engineering with rule injection in simulating phenomena such as disfluencies and interruptions, we expose the unreliability of LLM prompting. Our findings demonstrate that the proposed rule-based approach significantly outperforms prompt-based methods in both behavioral authenticity and controllability. Furthermore, this work emphasizes that simulator evaluations must audit actual generated behaviors rather than relying solely on configuration parameters, thereby establishing a robust paradigm for constructing high-fidelity spoken interaction evaluation systems.

0 citationsRead paper
Recent publications

Latest Papers

Back in Style: A Sociolinguistic Approach to Authoring and Measuring Persona Fidelity in User Simulation

Oct 07, 2026

This study addresses the low fidelity of user simulators and their reliance on costly, subjective LLM-based evaluation by proposing an objective assessment framework grounded in sociolinguistics. Methodologically, personality is reconceptualized as observable linguistic style rather than a label prediction task, enabling deterministic measurement through specific style rates. By integrating stylometric and lexicon-based analysis techniques, the framework introduces model-free validation tools for objective diagnostics. Experimental results demonstrate that this approach significantly enhances both style adherence and discriminability across most models, offering a novel pathway for precisely identifying simulation errors and generating diverse user profiles.

0 citationsRead paper

Prompts versus Rules: Auditing and Controlling Speech Naturalness Behaviors in Voice User Simulators

Oct 07, 2026

This study addresses the disconnect between configured naturalness behaviors and actual outputs in spoken user simulators, which compromises evaluation validity. To this end, we propose a linguistics-based deterministic rule injection algorithm alongside a naturalness behavior auditing framework. By systematically comparing large language model (LLM) prompt engineering with rule injection in simulating phenomena such as disfluencies and interruptions, we expose the unreliability of LLM prompting. Our findings demonstrate that the proposed rule-based approach significantly outperforms prompt-based methods in both behavioral authenticity and controllability. Furthermore, this work emphasizes that simulator evaluations must audit actual generated behaviors rather than relying solely on configuration parameters, thereby establishing a robust paradigm for constructing high-fidelity spoken interaction evaluation systems.

0 citationsRead paper