21 days ago
-
Add dataset card and per-directory licence note
-
Add backups: E19 pool-b and soft-label paid-output archives (no a2)
-
Add xover_* crossover synthetic datasets (BKT+IRT simulator)
-
Add synth_prior/f2_flat_s0 (E15 F2 flat prior corpus)
-
Add synth_prior/f2_wide_s0 (E15 F2 wide prior corpus) [4/4]
-
Add synth_prior/f2_wide_s0 (E15 F2 wide prior corpus) [3/4]
-
Add synth_prior/f2_wide_s0 (E15 F2 wide prior corpus) [2/4]
-
Add synth_prior/f2_wide_s0 (E15 F2 wide prior corpus) [1/4]
-
Add synth_prior/e18_s0 (E18 F2-prior corpus) [3/3]
-
Add synth_prior/e18_s0 (E18 F2-prior corpus) [2/3]
-
Add synth_prior/e18_s0 (E18 F2-prior corpus) [1/3]
-
Add synthetic item banks and E8c synthetic learners (synth_bank, synth_e17, synth_s, synth_l)
-
Add e20 MIX-E18E19 and TA arms (synthetic sources only)
-
Add e19 LLM-student simulator, pools a1 (GSM8K+MATH, MIT) and b (E17 bank); a2 excluded
-
Add e18 F2-prior grid arms (G-*) + task store + Jev teacher labels
-
Add e17 data-free recipes (Y1/Y1b/Y2/Y3/Y3G/Y4, G, _g) + Jev teacher labels for G/Y1/Y3
-
Add e16 CC0 / CC BY-SA sequence + forecast tasks (T-seq, T-fc, P-seq, D-seq) and replay (Apache-2.0)
-
Add e15 data-free arms (F2-flat, F2-syn wide prior; F3-gen/F4-mix from Apache-2.0 jev-distill corpus)
-
initial commit
