🎉 New preprint: Can #LLMs automate reproducibility checks?
👉We test this on 76 social & behavioral studies
✅Our LLM pipeline arrived at original qualitative conclusions in 96% of cases
🎯We recovered the effect sizes in 41% (human reanalysts: 34%)
arxiv.org/abs/2606.13670
Reproducibility in the social and behavioral sciences is typically evaluated by independent researchers who reanalyze the original data to assess whether the published findings can be recovered. Howev...
