Researchers have developed DEFINED, a computational framework designed to assess creativity in debate scenarios using a data-efficient approach. This system employs a pre-trained language model with a hierarchical scoring head to evaluate eight dimensions of creativity, distinguishing between divergent and convergent thinking. The framework was trained on authentic debate data and utilizes a mixed-granularity strategy to learn from limited expert supervision, outperforming existing automated scoring methods and prompt-based LLM evaluators. AI
IMPACT This framework offers a more efficient and accurate method for evaluating creativity in complex, open-ended tasks, potentially reducing reliance on costly human evaluation.
RANK_REASON The cluster contains an academic paper detailing a new computational framework for assessing creativity.
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →