time	model	prompt	layer	strength	n_tricks	n_plain	n_random	hold	hold_lo	hold_hi	switch	switch_lo	switch_hi	fluent	random_cjk	random_hi	score	score_lo	score_hi
2026-10-26T12:08:42	qwen8b	S4	14	1.75	24	15	45	0.958	0.875	1.0	0.831	0.79	0.864	1.0	0.0	0.0	0.796	0.714	0.856
2026-10-26T12:13:12	gemma2-9b	S3	24	4.0	24	15	45	0.875	0.75	1.0	0.783	0.744	0.819	0.933	0.0	0.0	0.639	0.495	0.763
2026-10-26T12:18:02	llama8b	S4	15	1.25	24	15	45	1.0	1.0	1.0	0.664	0.578	0.746	0.667	0.0	0.0	0.503	0.309	0.678
2026-10-26T12:19:48	gemma2-2b	S4	8	1.5	24	15	45	0.875	0.708	1.0	0.729	0.704	0.757	1.0	0.0	0.0	0.638	0.535	0.724
2026-10-26T12:24:42	llama8b	S2	15	1.25	24	15	45	1.0	1.0	1.0	0.734	0.671	0.791	0.867	0.0	0.0	0.666	0.514	0.784
2026-10-26T12:29:57	llama8b	S4	15	1.0	24	15	45	1.0	1.0	1.0	0.681	0.628	0.732	0.867	0.0	0.0	0.609	0.469	0.727
2026-10-26T12:34:45	llama8b	S2	15	1.0	24	15	45	1.0	1.0	1.0	0.647	0.561	0.718	0.8	0.0	0.0	0.553	0.399	0.691
2026-10-26T19:11:38	llama8b	S2	15	1.25	24	15	45	1.0	1.0	1.0	0.734	0.671	0.791	0.867	0.0	0.0	0.666	0.514	0.784