# Source index

These sources support methodology claims; they do not prescribe universal
sample sizes, agreement thresholds, pay rates, or stopping values. Recheck
versions and applicability before consequential use.

| Source | Use |
|---|---|
| Burr Settles, “Active Learning Literature Survey,” 2009, https://burrsettles.com/pub/settles.activelearning.pdf | Uncertainty, diversity, and query-strategy framing; not a deployment recipe. |
| Cohn, Atlas, and Ladner, “Improving generalization with active learning,” Machine Learning 1994, https://doi.org/10.1007/BF00993150 | Early pool-based active-learning evidence; transfer limits depend on task and sampling assumptions. |
| Artstein and Poesio, “Inter-Coder Agreement for Computational Linguistics,” Computational Linguistics 2008, https://doi.org/10.1162/coli.2008.34.4.555 | Agreement interpretation and the distinction between reproducibility and truth. |
| NIST AI RMF 1.0, https://doi.org/10.6028/NIST.AI.100-1 | Risk, governance, and documentation context; not an annotation threshold catalog. |