Rethinking Theory of Mind Benchmarks for LLMs: Towards A User-Centered Perspective
This work addresses the problem of inaccurate social intelligence assessment in LLMs for researchers and developers, but it is incremental as it builds on existing critiques without presenting new empirical results.
The paper critiques the use of human-designed Theory of Mind tasks to benchmark LLMs, highlighting theoretical and methodological limitations, and proposes a user-centered, dynamic approach to redefine ToM benchmarks based on user interactions and experiences.
The last couple of years have witnessed emerging research that appropriates Theory-of-Mind (ToM) tasks designed for humans to benchmark LLM's ToM capabilities as an indication of LLM's social intelligence. However, this approach has a number of limitations. Drawing on existing psychology and AI literature, we summarize the theoretical, methodological, and evaluation limitations by pointing out that certain issues are inherently present in the original ToM tasks used to evaluate human's ToM, which continues to persist and exacerbated when appropriated to benchmark LLM's ToM. Taking a human-computer interaction (HCI) perspective, these limitations prompt us to rethink the definition and criteria of ToM in ToM benchmarks in a more dynamic, interactional approach that accounts for user preferences, needs, and experiences with LLMs in such evaluations. We conclude by outlining potential opportunities and challenges towards this direction.