The dataset is currently empty. Upload or create new data files. Then, you will be able to explore them in the Dataset Viewer.
Insider LLM Detection Results (archived)
Status: archived. This repository holds no result files. It was created as the results mirror of the earlier Insider LLM Detection study (Aug–Sep 2026), but no results were uploaded before the study was frozen on 2026-09-25.
- Benchmark of the earlier study: INSIDER_LLM_DETECTION_BENCHMARK (archived)
- Code: INSIDER_LLM_DETECTION_CODE
- Pilot write-up (about 20 runs; pipeline validation, not study data): umaysamli.com/blog/trusting-models-to-log-their-own-actions
The work continues in four separate projects, listed in the benchmark card.
- Downloads last month
- 63