Remove superseded perturbed_queries_metadata.json (replaced by generation_config.json and seeds.json)

Remove superseded perturbed_queries/ JSON files (replaced by per-split TSVs in msmarco-dev/, trec-dl-2019/, trec-dl-2020/)

Split perturbations by eval split (msmarco-dev/trec-dl-2019/trec-dl-2020), add TSV format with original+perturbed query text, add data_download scripts for MS MARCO/TREC-DL originals

Add weight-file links and from_pretrained usage snippet

Correct provenance, licensing, and citation details: lost_in_decoding_artifacts/README.md

Correct provenance, licensing, and citation details: identifiers/README.md

Correct provenance, licensing, and citation details: LICENSE.md

Correct provenance, licensing, and citation details: README.md

Upload PAG checkpoint, document identifiers, and Lost-in-Decoding evaluation artifacts

initial commit