bursty_K6

done · peak unseen_mismatch 95.8

Referenced by

α 3.0 ×150 α 1.0 ×150 α 2.1429 ×150 α 1.0 ×150 α 1.2857 ×150 α 1.0 ×150

accuracy by training iteration, mean ± sd over seeds · schedule ends at α = 1.0

seen_* probed at reference α = 1.4 · unseen_* = held-out pairings, uniform (α-independent) · 1000 pairs per condition

iteration	seen_match	seen_mismatch	unseen_match	unseen_mismatch
25	51.1 ± 1.2	49.8 ± 2.0	49.8 ± 1.3	50.4 ± 1.4
50	51.5 ± 1.8	50.0 ± 1.7	50.5 ± 1.1	49.7 ± 1.5
75	52.2 ± 3.3	49.5 ± 1.6	50.9 ± 2.7	49.6 ± 2.7
100	55.6 ± 9.1	52.2 ± 4.7	53.4 ± 8.5	50.4 ± 3.5
125	63.7 ± 13.4	56.8 ± 10.7	61.0 ± 13.1	54.6 ± 9.8
150	78.9 ± 16.4	70.1 ± 14.2	74.9 ± 16.8	66.4 ± 13.4
175	91.2 ± 14.5	71.0 ± 15.7	90.3 ± 15.2	68.9 ± 14.7
200	97.0 ± 9.2	69.7 ± 19.5	96.8 ± 9.8	68.2 ± 19.3
225	99.8 ± 0.7	65.1 ± 20.2	99.8 ± 0.7	64.6 ± 19.2
250	100.0 ± 0.0	58.4 ± 26.8	100.0 ± 0.0	57.5 ± 26.4
275	100.0 ± 0.0	61.3 ± 27.0	100.0 ± 0.0	60.9 ± 26.6
300	100.0 ± 0.0	59.5 ± 31.1	100.0 ± 0.0	59.8 ± 31.0
325	100.0 ± 0.0	68.7 ± 29.8	99.9 ± 0.4	68.3 ± 29.4
350	99.9 ± 0.2	81.4 ± 22.3	99.6 ± 1.1	80.0 ± 22.0
375	99.7 ± 0.6	93.8 ± 11.8	99.4 ± 1.4	92.0 ± 12.3
400	99.7 ± 0.6	96.1 ± 11.0	98.5 ± 3.2	93.7 ± 11.6
450	99.3 ± 1.0	95.6 ± 10.9	97.7 ± 3.8	92.8 ± 11.5
500	100.0 ± 0.1	94.5 ± 12.7	99.8 ± 0.5	93.4 ± 13.3
550	100.0 ± 0.1	97.2 ± 9.5	99.8 ± 0.5	95.8 ± 10.0
600	100.0 ± 0.0	98.1 ± 2.8	99.9 ± 0.3	94.9 ± 7.7
650	100.0 ± 0.1	95.9 ± 10.7	99.2 ± 1.8	92.2 ± 11.9
700	99.9 ± 0.2	98.2 ± 2.0	99.3 ± 1.3	93.8 ± 8.1
750	99.9 ± 0.1	97.7 ± 3.4	99.2 ± 1.4	93.5 ± 7.3
800	100.0 ± 0.1	98.6 ± 2.8	99.8 ± 0.4	94.8 ± 7.7
850	100.0 ± 0.0	98.1 ± 2.8	99.8 ± 0.5	93.8 ± 8.1
900	100.0 ± 0.0	97.3 ± 4.4	99.5 ± 0.9	91.8 ± 10.4