Operation Veja: Fixing Fundamental Concepts Missing from Modern Roleplaying Training Paradigms

TL;DR

This paper introduces the VEJA framework, emphasizing the importance of Values, Experiences, Judgments, and Abilities in training roleplaying models to achieve more authentic and engaging characters, addressing limitations of current paradigms.

Contribution

The paper proposes the VEJA framework as a novel data curation paradigm that enhances character authenticity in roleplaying models, supported by a pilot study showing significant quality improvements.

Findings

VEJA-grounded dataset outperforms synthetic baselines in quality.

Shift to conceptually grounded data improves character depth.

Significant quality gap demonstrated by LLM-as-judge evaluation.

Abstract

Modern roleplaying models are increasingly sophisticated, yet they consistently struggle to capture the essence of believable, engaging characters. We argue this failure stems from training paradigms that overlook the dynamic interplay of a character's internal world. Current approaches, including Retrieval-Augmented Generation (RAG), fact-based priming, literature-based learning, and synthetic data generation, exhibit recurring limitations in modeling the deliberative, value-conflicted reasoning that defines human interaction. In this paper, we identify four core concepts essential for character authenticity: Values, Experiences, Judgments, and Abilities (VEJA). We propose the VEJA framework as a new paradigm for data curation that addresses these systemic limitations. To illustrate the qualitative ceiling enabled by our framework, we present a pilot study comparing a manually curated, VEJA-grounded dataset against a state-of-the-art synthetic baseline. Using an LLM-as-judge evaluation, our findings demonstrate a significant quality gap, suggesting that a shift toward conceptually grounded data curation, as embodied by VEJA, is necessary for creating roleplaying agents with genuine depth and narrative continuity. The full dataset is available at https://github.com/HyouinKyoumaIRL/Operation-Veja