Scientific reports
November 30, 2024
Takayuki Niizato, Kotaro Sakamoto, Yoh-Ichi Mototake et al.
9 citations
Integrated information theory (IIT), which measures consciousness via information integration, can be applied to non-biological systems. In schooling fish (Plecoglossus altivelis), group integrity (Φ) was highest at the critical state. Multiple levels of criticality existed as distinct subgroups within the school, and these fragmented critical subgroups coexisted with the group's overall criticality. The distribution of high-criticality subgroups was uneven across time and space. Fish in high-criticality subgroups were less affected by internal and external stimuli than those in low-criticality subgroups. These results align with prior interpretations of critical phenomena and provide a new view of empirical critical-state dynamics.
Scientific reports
October 26, 2022
Takayuki Niizato, Yuta Nishiyama, Kotaro Sakamoto et al.
During the rubber hand illusion, the feeling of ambiguous body ownership arises from a conflict between what one feels and what one judges. Integrated information theory (IIT) measures the difference between the integrity of a whole system and its subsystems. Analyzing seven time-series of physiological data from a small body-brain system, the authors demonstrate that the integrity of the whole system decreases during the illusion while the integrity of its subsystems increases. Subjective ratings of ambiguous body ownership were associated with the measure of integrated information. The results suggest that IIT can explain both the general tendency of ownership illusions and individual differences in subjective experience.
Entropy (Basel, Switzerland)
June 30, 2020
Takayuki Niizato, Kotaro Sakamoto, Yoh-Ichi Mototake et al.
Integrated information theory (IIT) 3.0, which measures a system's intrinsic cause-effect structure, can classify fish schools by group size and integrity when applied to collective behavior in Plecoglossus altivelis. Using global parameter settings and multiple timescales (from 5/120 to 120/120 seconds), the analysis successfully distinguished schools of different sizes and degrees of group integrity near the fish's reaction time scale. At longer timescales, interaction heterogeneity diminished compared to shorter timescales. The findings also offer two tentative answers to the heap paradox, a longstanding puzzle in collective behavior about how individual actions combine into group-level properties.
PloS one
January 1, 2020
Takayuki Niizato, Kotaro Sakamoto, Yoh-Ichi Mototake et al.
Integrated information theory (IIT), originally developed to measure consciousness, can reveal a new type of leadership in fish schools that other measures cannot detect. Applying IIT 3.0 to real fish schools (Plecoglossus altivelis) showed a discontinuity in the average integrated information 〈Φ(N)〉 for schools of four or more fish, a transition not observed with mutual information or transfer entropy. This discontinuity correlated with the emergence of leadership, defined not by group behavior but by autonomy or group integrity. Simulations with the Boids model produced different results from real fish, suggesting that IIT captures a unique aspect of collective dynamics.
arXiv Preprint Archive
December 3, 2018
Takayuki Niizato, Kotaro Sakamoto, Yoh-Ichi Mototake et al.
Integrated Information Theory (IIT), originally developed to measure consciousness, can reveal emergent properties in fish schools that other measures miss. Applying IIT 3.0 to real ayu fish (Plecoglossus altivelis) schools, the authors found that leadership behavior emerges only when school size reaches four or more individuals, not at sizes of three or fewer. This discontinuous transition was not detected by mutual information or in a simple Boids model. The results suggest that integrated information Φ captures inherent collective properties invisible to other metrics, and that figure-ground perception in fish may relate to optimal paths for the functional evolution of collective behavior.