Physical mechanism of mind changes and tradeoffs among speed, accuracy, and energy cost in brain decision making: Landscape, flux, and path perspectives
Yan Han1, 2, Zhang Kun2, Wang Jin1, 2, 3, †,
       

The potential landscapes for different stimulus inputs and motion coherence. (a)–(c) Potential landscapes in the state space of average gating variables S1 and S2 at the zero coherence level for different stimulus inputs, where the stimuli strength μ0 = 0, 10 and 30 Hz, respectively. The central attractor corresponding to the undecided state is indicated by c in the figure, and a and b on two sides indicate the two decided states. (d)–(f) Two-dimensional potential landscapes at the non-zero coherence level, where μ0 = 30 Hz and the coherence c′ = 0.2, 0.3 and 0.65, respectively. In all these subgraphs, the parameters a = 269.5 and b = 108 and the diffusion coefficient D = 3.6 × 10−7.