Abstract:We investigate the metric mean dimension of subshifts of compact type. We prove that the metric mean dimensions of a continuous map and its inverse limit coincide, generalizing Bowen's entropy formula. Building upon this result, we extend the notion of metric mean dimension to discontinuous maps in terms of suitable subshifts. As an application, we show that the metric mean dimension of the Gauss map and that of induced maps of the Manneville-Pomeau family is equal to the box dimension of the corresponding set of discontinuity points, which also coincides with a critical parameter of the pressure operator associated to the geometric potential.
What problem does this paper attempt to address?
The core problem that this paper attempts to solve is how to study the metric mean dimension through subshifts of compact type. Specifically, the author explores the metric mean dimension of subshifts of compact type and proves that the metric mean dimensions of a continuous mapping and its inverse limit are equal, generalizing Bowen's entropy formula. Based on this result, the author extends the concept of metric mean dimension to discontinuous mappings and defines it through appropriate subshifts of compact type.
In addition, the author applies these results to show that the metric mean dimensions of the Gauss mapping and the Manneville - Pomeau family - induced mapping are equal to the box dimension of the corresponding sets of discontinuous points, which also coincides with the key parameters of the pressure operator associated with the geometric potential.
### Main contributions:
1. **Generalize Bowen's entropy formula**: Prove that the metric mean dimensions of a continuous mapping and its inverse limit are equal.
2. **Extension to discontinuous mappings**: Propose a definition of metric mean dimension applicable to discontinuous mappings.
3. **Application examples**: Show the relationship between the metric mean dimension of specific mappings (such as the Gauss mapping and the Manneville - Pomeau family mapping) and the box dimension of their sets of discontinuous points.
### Key theorems:
- **Theorem A**: For a compact metric space \((X, d)\) and a closed subset \(\Gamma\subset X\times X\),
\[
\text{mdim}_M(\Gamma^{\mathbb{N}}, d^{\mathbb{N}}, \sigma^{\mathbb{N}})=\text{mdim}_M(\Gamma^{\mathbb{Z}}, d^{\mathbb{Z}}, \sigma^{\mathbb{Z}})
\]
- **Theorem B**: For a compact metric space \((X, d)\) and a continuous mapping \(T:X\rightarrow X\),
\[
\text{mdim}_M(X, d, T)=\text{mdim}_M(\Gamma^{\mathbb{N}}, d^{\mathbb{N}}, \sigma^{\mathbb{N}})=\text{mdim}_M(\Gamma^{\mathbb{Z}}, d^{\mathbb{Z}}, \sigma^{\mathbb{Z}})
\]
- **Theorem C**: For a compact interval \(I\subset\mathbb{R}\) and a zero - measure compact set \(A\subset I\),
\[
\text{mdim}_M(I, d, T_A)=\dim_B(A, d)
\]
These results not only deepen the understanding of the metric mean dimension but also provide a new perspective for dealing with discontinuous mappings.