On the de Rham flip-flopping in dual towers

This paper establishes a de Rham and Hyodo-Kato flip-flopping theorem for dual towers of rigid analytic spaces by utilizing pro-étale comparison theorems, and applies this result to prove the admissibility of the de Rham and Hyodo-Kato cohomologies of finite level coverings of Drinfeld spaces as representations of $\mathbb{GL}_{d+1}(K)$.

The Gist

Here is an explanation of the paper "On de Rham Flip-Flopping in Dual Towers" using simple language, analogies, and metaphors.

The Big Picture: Two Sides of the Same Coin

Imagine you have a complex, multi-layered mathematical structure called a "Tower." In the world of number theory, these towers are built by stacking layers of geometric shapes (like spheres or donuts) on top of each other, getting more and more detailed as you go up.

There are two famous types of these towers:

1. The Drinfeld Tower: Think of this as a tower built from "Drinfeld space," which is a specific kind of geometric landscape.
2. The Lubin-Tate Tower: Think of this as a tower built from "Lubin-Tate space," a different landscape that looks very different on the surface.

For a long time, mathematicians knew that if you looked at these towers through a specific "lens" (called $\ell$-adic cohomology), they were actually identical. It was like looking at a cat and a dog through a special filter that made them both look like perfect spheres. You could swap them, and the math worked out perfectly. This is called "Flip-Flopping."

The Problem:
However, when mathematicians tried to look at these towers through a different, more powerful lens (called de Rham and Hyodo-Kato cohomology), the magic stopped working.

• The "Drinfeld" tower and the "Lubin-Tate" tower looked completely different.
• The tools used to prove they were the same before (perfectoid spaces) didn't work for this new lens because the "differential forms" (the mathematical equivalent of measuring slopes and curves) on the base of the tower didn't match up with the top of the tower.

The Goal:
The authors, Gabriel Dospinescu and Wiesława Nizioł, wanted to prove that even with this new, difficult lens, the two towers are still secretly the same. They wanted to show that you can still "flip-flop" between them.

---

The Solution: The "Universal Translator"

To solve this, the authors didn't try to compare the towers directly. Instead, they built a Universal Translator.

1. The "Diamond" (The Perfectoid Space)

Imagine the two towers are two different languages. You can't translate them word-for-word easily. But, if you go to the very top of both towers (the "infinity" level), they merge into a single, perfect, infinite object called a Perfectoid Space (or a "Diamond").

• Analogy: Think of the Drinfeld and Lubin-Tate towers as two different dialects of a language. At the very top of the mountain, they merge into a single, pure "Universal Language."

2. The "Period Sheaves" (The Dictionary)

The authors realized that while the towers look different, they both share a common underlying structure when viewed through a specific mathematical dictionary called Period Sheaves (specifically $B_{dR}$ and $B$).

• Analogy: Imagine the towers are two different buildings. You can't walk from one to the other directly because the doors don't align. But, both buildings are built on the same foundation (the Diamond). The authors found a special set of blueprints (the Period Sheaves) that describe the foundation perfectly.
• Because these blueprints are built on the foundation, they automatically work for both buildings. They "descend" down to both towers naturally.

3. The "Flip-Flop" Mechanism

The authors used a clever trick:

1. They translated the "Drinfeld" tower into the language of the Universal Foundation (using the Period Sheaves).
2. They translated the "Lubin-Tate" tower into the same language.
3. Since the foundation is the same, the translations matched perfectly!
4. They then proved that you could translate back down to the original towers, effectively swapping them.

The Result: They proved that the "de Rham" and "Hyodo-Kato" cohomologies (the measurements of the towers) are isomorphic. In plain English: The Drinfeld tower and the Lubin-Tate tower are mathematically identical, even when you measure them with the most sensitive tools available.

---

Why Does This Matter? (The "Admissibility" Bonus)

The paper doesn't just say "they are the same." It also proves something very useful about symmetry.

• The Group Action: These towers have groups of symmetries acting on them (like rotating a cube). In the Drinfeld tower, the group is $GL_{d+1}(K)$ (a group of matrices). In the Lubin-Tate tower, it's $D^\times$ (units in a division algebra).
• The Discovery: The authors showed that the "shape" of the data on the Drinfeld tower is exactly the same as the "shape" of the data on the Lubin-Tate tower, even when you consider how these symmetry groups twist and turn the data.
• The Metaphor: Imagine two different orchestras playing the same symphony. One orchestra uses violins (Drinfeld), and the other uses flutes (Lubin-Tate). The authors proved that even though the instruments are different, the music (the representation theory) is identical. Furthermore, they proved that this music is "admissible," meaning it's well-behaved and finite enough to be studied and understood by mathematicians.

Summary of the "Magic Trick"

1. The Problem: Two towers look different when viewed with advanced mathematical tools.
2. The Insight: They share a common "perfect" foundation at infinity.
3. The Tool: The authors used "Period Sheaves" (special mathematical blueprints) that live on this foundation.
4. The Trick: Because the blueprints live on the shared foundation, they automatically work for both towers. This allows the authors to translate the math from one tower to the other seamlessly.
5. The Payoff: They proved the towers are identical in a deep, structural way, and that the symmetries acting on them are also identical.

In a nutshell: The authors built a bridge between two seemingly different mathematical worlds using a shared foundation, proving that they are actually two sides of the same coin, even under the most rigorous scrutiny.

Technical Summary

Here is a detailed technical summary of the paper "On de Rham Flip-Flopping in Dual Towers" by Gabriel Dospinescu and Wiesława Nizioł.

1. Problem Statement and Motivation

The paper addresses a fundamental asymmetry in the comparison of cohomology theories for dual towers of rigid analytic spaces, specifically the Drinfeld tower and the Lubin-Tate tower (and their generalizations to local Shimura varieties).

• The Context: It is a classical result (Faltings, Fargues, Scholze) that for $\ell \neq p$, the $\ell$-adic cohomology with compact support of the Lubin-Tate tower is isomorphic to that of the Drinfeld tower. This is deduced from the fact that both towers are "decompletions" of a single perfectoid space.
• The Failure of $p$-adic Theories: While this "flip-flopping" holds for $p$-torsion étale cohomology, it fails dramatically for $p$-adic étale or pro-étale cohomology.
• The Gap: A previous result by the authors (in [10]) established a "Hodge flip-flopping" isomorphism for the compactly supported de Rham cohomology of the 1-dimensional Drinfeld and Lubin-Tate towers. However, this relied on arguments involving $p$-adic period maps that did not generalize easily to higher dimensions.
• The Core Question: Can one establish an isomorphism between the compactly supported de Rham and Hyodo-Kato cohomologies of dual towers in arbitrary dimensions, and can this be extended to general dual towers of local Shimura varieties?

2. Methodology and Strategy

The authors develop a new strategy that avoids the limitations of previous period map arguments by utilizing condensed mathematics and relative period sheaves.

A. The "Flip-Flopping" Mechanism

The central idea is to express the cohomologies of interest (de Rham and Hyodo-Kato) in terms of pro-étale cohomology of relative period sheaves.

1. Period Sheaves as Avatars: Instead of working directly with differential forms on the base spaces (which do not descend well to the infinite level), the authors use period sheaves ($B_{dR}$, $B_{pdR}$, $B$, $B_{pFF}$) defined on the pro-étale site.
2. Descent Property: These period sheaves satisfy pro-étale descent. Consequently, their pro-étale cohomology on the "completed" infinite tower (which is a perfectoid space) naturally descends to both dual towers. This creates a natural isomorphism (flip-flop) at the level of period sheaf cohomology.
3. Comparison Theorems: The authors utilize recent comparison theorems (Bosco, Colmez-Gilles-Nizioł) that relate the cohomology of period sheaves to the classical de Rham and Hyodo-Kato cohomologies, twisted by the period rings.
   • De Rham: $R\Gamma_{dR}(X) \simeq R\Gamma(G_K, R\Gamma_{pro\acute{e}t}(X_C, B_{pdR}))$.
   • Hyodo-Kato: $R\Gamma_{HK}(X) \simeq R\Gamma(G_K, R\Gamma_{pro\acute{e}t}(X_C, B))$.
4. Removing the Twist: The comparison isomorphisms involve twists by period rings (e.g., $B_{dR}$). The authors show that taking Galois fixed points (or derived fixed points) on the cohomology level effectively "kills" these twists, transferring the flip-flop isomorphism from the period sheaf level to the de Rham/Hyodo-Kato level.

B. Technical Framework: Condensed Mathematics

The paper heavily relies on the framework of condensed and solid abelian groups (developed by Clausen and Scholze).

• Solid Modules: Cohomology groups are treated as objects in the category of solid modules over $K$ or $\breve{C}$. This allows for the handling of infinite-dimensional representations and topological vector spaces (Fréchet spaces) in a robust categorical setting.
• Smooth Representations: The authors prove that the group actions on these cohomology groups are smooth (vectors are fixed by open subgroups), which is crucial for applying representation-theoretic cancellation arguments.

C. Representation Theoretic Input

To pass from isomorphisms of derived objects (quasi-isomorphisms) to isomorphisms of specific cohomology groups (and to remove Lie algebra cohomology twists), the authors employ:

• Künneth Formulas: Relating the cohomology of a product to the tensor product of cohomologies.
• Cancellation Theorems: Using the Krull-Schmidt theorem and properties of admissible representations of reductive groups (specifically, if $\pi_1 \oplus \pi \cong \pi_2 \oplus \pi$ and representations are admissible, then $\pi_1 \cong \pi_2$).
• Lie Algebra Cohomology: They show that for compact $p$-adic Lie groups, the Lie algebra cohomology groups $H^i(\mathfrak{g}, K)$ have dimensions matching those of the dual group, allowing the "untwisting" of the isomorphisms.

3. Key Contributions and Results

A. Generalized Flip-Flopping Theorems

The paper proves two main theorems establishing isomorphisms for dual towers:

1. Theorem 1.1 (Drinfeld and Lubin-Tate):
   For the Drinfeld tower ($M^\varpi_\infty$) and Lubin-Tate tower ($LT^\varpi_\infty$) of dimension $d-1$ over $K$:
   $$H^i_{dR,c}(M^\varpi_\infty, C) \simeq H^i_{dR,c}(LT^\varpi_\infty, C)$$
   $$H^i_{HK,c}(M^\varpi_\infty, C) \simeq H^i_{HK,c}(LT^\varpi_\infty, C)$$
   These isomorphisms are equivariant under $GL_{d+1}(K) \times D^\times$ and hold in the category of solid modules.

2. Theorem 1.2 (Local Shimura Varieties):
   The result generalizes to any dual tower of basic local Shimura varieties $(Sh(G, b, \mu)_\infty, Sh(\check{G}, \check{b}, \check{\mu})_\infty)$. There exist $G \times \check{G}$-equivariant isomorphisms for de Rham and Hyodo-Kato cohomologies.

B. Admissibility of Representations

A major application of these results is the proof of admissibility.

• The authors show that the compactly supported de Rham and Hyodo-Kato cohomologies of finite level coverings of the Drinfeld space (and general local Shimura varieties) are admissible representations of the associated $p$-adic Lie groups (e.g., $GL_{d+1}(K)$).
• This is achieved by transferring the known admissibility from the Lubin-Tate side (where it was previously established or easier to prove) to the Drinfeld side via the flip-flop isomorphism.

C. Lie Algebra and Hodge Flip-Flopping

The paper derives corollaries regarding:

• Lie Algebra Flip-Flopping: Isomorphisms involving Lie algebra cohomology $R\Gamma(\mathfrak{g}, -)$, which allow the removal of twists to get direct isomorphisms between cohomology groups at finite levels.
• Hodge Flip-Flopping: Isomorphisms for the cohomology of differential forms $\Omega^j$, derived from the filtered structure of the de Rham comparison.

4. Significance

• Unification of $p$-adic Hodge Theory: The paper successfully extends the "flip-flopping" phenomenon, previously known only for $\ell$-adic cohomology and 1-dimensional Hodge cohomology, to the full $p$-adic de Rham and Hyodo-Kato cohomologies in arbitrary dimensions.
• New Tools for Local Langlands: By establishing the admissibility of these cohomologies as representations of $p$-adic groups, the results provide a robust foundation for the Local Langlands Correspondence in the $p$-adic setting. It confirms that the geometric cohomology of these towers behaves well under the action of the relevant groups.
• Methodological Advancement: The shift from using explicit period maps (which are hard to generalize) to using period sheaves and pro-étale descent in the condensed setting offers a powerful new toolkit for studying $p$-adic cohomology of rigid analytic spaces. This approach bypasses the difficulties of defining differential forms on perfectoid spaces directly.
• Generalization: The results apply not just to the classical Drinfeld/Lubin-Tate case but to the broad class of local Shimura varieties, suggesting a deep structural symmetry in the $p$-adic geometry of these spaces that was previously inaccessible.

In summary, Dospinescu and Nizioł provide a rigorous, categorical proof that the de Rham and Hyodo-Kato cohomologies of dual towers are isomorphic, leveraging the power of condensed mathematics to resolve long-standing obstructions in $p$-adic Hodge theory.