Table 2. Linear mixed effects model fit (adjusted R²) and standard deviation (SD) of random effects.

	R2	SD
Random effect		+ (1|f)#	+ (1|f:n)^	+ (1|n)&
Φ ~ c	.489	0.011	3.185 × 10−11
Φ ~ c + l + d + l:c + d:c	.493	0.011	1.524 × 10−11
φ ~ c + s + c:s	.412	5.95 × 10−3	4.07 × 10−3
Δφ ~ s	.372	0.235	0.121
aW ~ F	.476	0.019
aA ~ F	.309	0.020
aΦW ~ l + d	.562			0.683
aΦA ~ l + d	.513			0.702
aφW ~ l + d	.555			0.686
aφA ~ l + d	.535			0.694

Model specifications are described in detail in Methods. Φ: system-level integrated information. c: level of arousal (wake/anesthesia). l: channel set location. d: sum of pairwise distances between channels within a channel set. φ: (mechanism-level) integrated information. s: mechanism size. Δφ: ratio of wakeful to anesthetized integrated information for mechanism-level integrated information. a_W: within-fly classification accuracy. F: feature used for classification (categorical variable; individual 1-, 2-, 3-, 4-channel mechanisms, 1 feature; IIS, 15 features; or system-level integrated information, 1 feature). a_A: across-fly classification accuracy. a_ΦW: within-fly classification accuracy using system-level integrated information. a_ΦA: across-fly classification using system-level integrated information. a_φW: within-fly classification accuracy using the IIS. a_φA: across-fly classification accuracy using the IIS.

# Random intercept for effect of fly.

^ Random intercept for interaction between fly and channel set.

& Random intercept for channel set.