Upload outputs/sample_complexity.csv with huggingface_hub

Upload outputs/figure1_stoptime.csv with huggingface_hub

Upload outputs/table1_runtime.csv with huggingface_hub

Upload outputs/verdict.json with huggingface_hub

Upload repro/src/run_suite.py with huggingface_hub

Upload repro/src/mnl_bsi.py with huggingface_hub

Upload pages/claim-6-figure-1/page.md with huggingface_hub

Upload pages/claim-5-table-1/page.md with huggingface_hub

Upload pages/claim-4-corollary-4-5/page.md with huggingface_hub

Upload pages/claim-3-theorem-4-4/page.md with huggingface_hub

Upload pages/claim-2-theorem-3-5/page.md with huggingface_hub

Upload pages/claim-1-theorem-3-3/page.md with huggingface_hub

Upload pages/conclusion/page.md with huggingface_hub

Upload pages/overview/page.md with huggingface_hub

Upload pages/index.md with huggingface_hub

Upload logbook.json with huggingface_hub

Claims [0][1][3][4]: implement Thm 3.3 exact MILP LMO (scipy.milp), Thm 3.5 lifted Frank-Wolfe (poly-time lifted LMO only), Cor 4.5 lifted-vs-exact (1+eps_lift) comparison, Table 1 MILP-vs-lifted cost

Update logbook: Repro - Optimal Design for Multinomial Logit Model

initial commit