Phase 1 — Motivation & Literature
Establish a theoretical foundation for EQ benchmarking through literature synthesis. Phase 1 involved a detailed evaluation of previous benchmarking frameworks alongside a comprehensive theoretical survey of emotional intelligence in both human and AI systems. These interdisciplinary efforts produced a first manuscript—now under peer review—that consolidates lessons from prior benchmarking frameworks and informs the benchmark framework that will guide the next phases of work. Constructionist accounts of emotion have implications for evaluation .
Goal
- Establish a theoretical foundation for EQ benchmarking through literature synthesis.
- Produce a manuscript that consolidates lessons from prior benchmarking frameworks and informs future phases of work.
Methodology
- Systematic reviews across philosophy, psychology, neuroscience, and computer science.
- Comparative analysis of existing EQ benchmarking frameworks and methodologies.
- Database creation detailing axes, definitions, measurement methods, and Sense, Explain, Respond (SER) axis alignment.
This stage focused on surveying multidisciplinary research and benchmarking practices to build a foundation for later phases. Phase 2 translates this outline into a reproducible, text-only, non-branching pilot over the SER axes.