| Name | Last commit | Updated |
|---|
|
artifacts
|
STATUS: Medium corpus v0, tokenizer, pretraining |
|
|
configs
|
Add arXiv/LaTeX, Crossref/PDF and books adapters (agents), governance docs,... |
|
|
data
|
Medium corpus v1 (14 470 docs / 451 072 eligible passages) summaries |
|
|
demo
|
Add arXiv/LaTeX, Crossref/PDF and books adapters (agents), governance docs,... |
|
|
docs
|
D19: mask same-passage negatives in InfoNCE |
|
|
model_cards
|
Add arXiv/LaTeX, Crossref/PDF and books adapters (agents), governance docs,... |
|
|
paper
|
Paper: non-results sections drafted (agent), amendment A1 (common W1 scale), doc... |
|
|
scripts
|
build_medium_corpus: configurable fixed splits |
|
|
src
|
D19: mask same-passage negatives in InfoNCE |
|
|
tests
|
Evaluation audits (identifiability, memorisation, Holm), MPS memory fixes (shape... |
|
|
.gitignore
|
Medium corpus v0 (13 372 docs / 371 419 eligible passages) summaries; tokenizer... |
|
|
CLAUDE.md
|
Add arXiv/LaTeX, Crossref/PDF and books adapters (agents), governance docs,... |
|
|
pyproject.toml
|
WP-AI-0002: repository scaffold, licence governance, PMC S3 ingest, JATS parser,... |
|
|
README.md
|
Add arXiv/LaTeX, Crossref/PDF and books adapters (agents), governance docs,... |
|