Entropy production rates for different notions of partial information

Abstract Experimentally monitoring the dynamics of a physical system, one cannot possibly resolve all microstates or transitions between them. Theoretically, these partially observed systems are modeled by considering only states and while rest hidden, merging into single mesostate, decimating unobserved states. The deviation system from thermal equilibrium can be characterized non-zero value entropy production rate (EPR). Based on information transitions, infer lower bound total EPR. Previous studies focused several approaches to optimize bounds EPR, fluctuation theorems associated with apparent regarding network topology inferred partial information, etc. Here, we calculate EPR values Markov chains driven external forces different notions information. We state-based coarse-graining, namely decimation two lumping protocols constraints, either preserving transition flux, occupancy number correlation function. Finally, compare cycle affinity. Our results further extended other networks various driving forces.