Investigation 47

Dataset: potions
Model: opus40
Training model: claude-opus-4-20250514
Inference model: gpt-4.1-mini
Example count: 3
Patience: 3
Current round: 576
Best round: 148

Rounds

Round	UUID	Started	Completed	Train Acc	Train KT	Val Acc	Val KT	Min Train/Val	Test Acc	Test KT
116	deacd5bf-0cdc-48b6-a8e9-1b02c367b77e	2025-06-24	2025-06-28 15:42:46.667139+00:00	0.600	0.222	0.400	0.397	0.400	0.550	0.260
138	448945af-ba03-437e-963e-bd2fa32d7b09	2025-06-28	2025-06-28 15:51:10.606845+00:00	0.613	0.213	0.650	0.188	0.613	0.450	0.347
139	943857e4-b9b4-4da6-9e05-b0be810fd458	2025-06-28	2025-06-28 16:00:57.561660+00:00	0.519	0.285	0.600	0.222	0.519	0.400	0.397
140	3607165c-306d-4774-8ea0-03249e515bf0	2025-06-28	2025-06-28 16:20:06.236662+00:00	0.531	0.275	0.700	0.156	0.531	0.350	0.455
141	eafb1e13-6294-42fe-b1cd-595c5d906043	2025-06-28	2025-06-28 16:40:44.822529+00:00	0.581	0.236	0.700	0.156	0.581	0.400	0.397
142	70427a3d-1c79-4f3d-b083-8f8f1c22af93	2025-06-28	2025-06-28 16:54:16.637964+00:00	0.662	0.179	0.750	0.126	0.662	0.500	0.301
143	41ade800-212b-4234-8647-21d52e163ed1	2025-06-28	2025-06-28 17:08:22.187525+00:00	0.700	0.156	0.800	0.098	0.700	0.500	0.301
144	2be102f6-71ff-4d48-9f94-019254c1560c	2025-06-28	2025-06-28 17:22:05.605865+00:00	0.713	0.148	0.850	0.071	0.713	0.450	0.347
145	f9ac58a2-0115-45f3-8afb-c12dad7af00b	2025-06-28	2025-06-28 17:40:43.651449+00:00	0.694	0.159	0.700	0.156	0.694	0.450	0.347
146	a7948fe5-af84-41eb-9d20-bc1c998d333d	2025-06-28	2025-06-28 17:58:48.996955+00:00	0.725	0.140	0.800	0.098	0.725	0.500	0.301
147	dcd6fea7-af26-4813-a18c-e2d5bf722a86	2025-06-28	2025-06-28 18:16:32.215767+00:00	0.750	0.126	0.700	0.156	0.700	0.500	0.301
148	4577883b-f828-4264-8780-4500b544e0ec	2025-06-28	2025-08-02 08:53:24.511769+00:00	0.738	0.133	0.850	0.071	0.738	0.450	0.347
556	7e8bd03b-df20-451f-b726-3de32741183d	2025-08-02	2025-08-03 10:53:24.576893+00:00	0.738	0.133	0.650	0.188	0.650	0.550	0.260
574	104b392d-e4e3-46ac-841a-a56d426e8905	2025-08-03	2025-08-03 10:57:18.840944+00:00	0.719	0.144	0.600	0.222	0.600	0.600	0.222
575	80e65d04-603f-4b70-9fa1-df9948a1d0ea	2025-08-03	2025-08-03 11:15:21.825934+00:00	0.706	0.152	0.550	0.260	0.550	0.600	0.222

Scores

round scores