Reproduce “TACT: A Pragmatic Approach to Qualitative Analysis with an AI Council” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Rudrajit Choudhuri, Christian Bird, Carmen Badea, Anita Sarma
The authors introduce TACT, a workflow that delegates qualitative response discovery and coding to independent AI models while keeping researchers responsible for interpretation. Software checks trace cited evidence to respondents’ words, and researchers review codes. On 192 Stack Overflow threads, TACT-Full achieved an F1 of 0.72, versus 0.65 for a single-model baseline, a result bounded to this forum evaluation.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.