CoDA builds qualitative codebooks in under 2 hours with 5x fewer examples
BrihiJ · x · 2026-10-25
CoDA is a human-LM collaborative tool that lets domain experts turn unstructured text into insights without weeks of annotation guideline writing: it develops qualitative-research codebooks in under 2 hours, using 5× fewer examples than manual codebook development.
Links to both the tool and the paper are provided — relevant for researchers and teams building labeling schemes.
More from Research
- The upside of AI-assisted research: failing faster at a higher level — mboehme_ · 2026-10-25
- Sparse MoE residual activations beat dense on interpretability, reports collab — amplifiedamp · 2026-10-25
- Can AlphaGenome-style models cure cancer? A debate on AI-driven small molecule drug design — Sauers_ · 2026-10-25
- Mathematicians debate whether AI-assisted research kills or accelerates insight — mboehme_ · 2026-10-25
- KLD Is Misunderstood: Why It Can't Rank Quantized Models Reliably — Positive-Resource922 · 2026-10-25
- Interpretability meme ranks the field from linear probes to 'Sacred Geometry' oracle mysticism — voooooogel · 2026-10-25