Atahan Karagoz
1 indexed paper
Recent (6 mo)
1With code
0Influential cites
0Benchmarked
0Publications per year
126
Top categories
AI×1NLP×1ML×1
Research Timeline
2026
A Persona-Based Evaluation Framework for Pluralistic Alignment in Generative AI
The paper proposes a persona-based evaluation framework that replaces monolithic AI benchmarks with structured cognitive profiles to capture diverse human perspectives, while also identifying the challenge of maintaining persona coherence over time.
Highlighted terms show continued research focus across papers