A Remark on Higher Homotopy Sheaves of Derived Arc Spaces

This paper extends Gaitsgory and Rozenblyum's result on derived arc spaces by demonstrating that, similar to smooth schemes, the derived arc spaces of reduced local complete intersection schemes are equivalent to their classical counterparts.

The Gist

The Big Picture: Smooth vs. Bumpy Roads

Imagine you are driving a car.

• Smooth Schemes (Smooth Schemes): These are like a perfectly paved, flat highway. If you drive on them, everything is predictable.
• Singular Schemes (Singular Schemes): These are like a road full of potholes, sharp turns, and bumps. In classical geometry, studying these roads is messy and complicated.

In recent years, mathematicians developed a new tool called Derived Algebraic Geometry. Think of this as a "super-microscope" or a "high-definition camera" that doesn't just look at the road, but looks at the texture of the road, the dust, and the microscopic vibrations.

Mathematicians Gaitsgory and Rozenblyum used this super-microscope to study Arc Spaces.

• What is an Arc Space? Imagine you are tracing a path (an "arc") on your road. An arc space is the collection of all possible paths you could draw on that road.
• The Discovery: They found that for smooth highways, the "super-microscope" view (Derived) looks exactly the same as the "naked eye" view (Classical). However, for bumpy roads (singularities), they suspected the two views might be very different. They thought the "super-microscope" would reveal hidden, complex structures (called higher homotopy sheaves) that the naked eye couldn't see.

The Author's Mission: The Reality Check

The author of this paper, Emile Bouaziz, decided to test this suspicion. He wanted to see if these hidden structures really exist for a specific, very common type of bumpy road: Reduced Local Complete Intersection (l.c.i.) schemes.

Think of an "l.c.i. scheme" as a road that is bumpy, but the bumps are "clean." They aren't random, chaotic messes; they are formed by the intersection of clean, well-defined curves (like the corner where two walls meet).

The Hypothesis: "Maybe for these 'clean' bumps, the super-microscope will still see something new and exciting."
The Result: "Actually, no. For these specific types of roads, the super-microscope sees exactly the same thing as the naked eye."

The Analogy: The "Ghost" in the Machine

To understand why this matters, imagine you have a sculpture made of clay.

1. Classical View: You look at the sculpture. You see the shape.
2. Derived View: You use a special scanner that detects "ghosts"—invisible layers of information that might exist underneath the clay.

For a smooth sphere, the scanner says, "No ghosts here."
For a jagged, broken rock, the scanner usually screams, "GHOSTS! There are hidden layers!"

Bouaziz looked at a specific type of broken rock (the l.c.i. scheme) and ran the scanner.
The Surprise: The scanner said, "No ghosts here either." The "Derived" version is identical to the "Classical" version.

How Did He Prove It? (The "Deformation" Trick)

The proof is a bit technical, but here is the metaphor:

Imagine you have a complex, 3D puzzle (the Derived Arc Space). You want to know if it's actually just a flat 2D drawing (the Classical Arc Space) pretending to be 3D.

Bouaziz built a time-lapse movie of the puzzle.

1. He started with the puzzle at time $t=0$.
2. He slowly "deformed" or morphed it over time.
3. He showed that as you watch the movie, the complex 3D structure smoothly collapses into the simple 2D drawing without tearing or breaking.

Because the complex structure can be smoothly morphed into the simple one, they are essentially the same thing. The "extra" 3D depth was an illusion.

Why is this "Disappointing" (and why that's good)?

The author admits in the paper that this result is "ultimately disappointing."

• Why? Because he hoped to find these hidden "ghosts" (invariants) that would help mathematicians solve difficult problems about singularities (the bumpy roads). He wanted to use the Derived view to unlock new secrets about how these shapes behave.
• The Reality: Since the Derived view is identical to the Classical view for these shapes, you don't get any new secrets from using the super-microscope.

However, this is actually a very important discovery.
It tells mathematicians: "Stop looking for magic here. For these specific shapes, the classical math is already perfect. You don't need the complicated Derived machinery to understand them." It saves everyone time and effort by telling them where not to look for new phenomena.

Summary in One Sentence

The paper proves that for a large, important class of "bumpy" geometric shapes, the fancy new "super-microscope" of modern mathematics sees absolutely nothing different than what we could already see with our naked eyes, meaning the complex hidden structures we hoped to find simply aren't there.

Technical Summary

1. Problem Statement and Context

The paper addresses a question arising from Derived Algebraic Geometry (DAG), specifically concerning the relationship between classical arc spaces and their derived counterparts.

• Background: In classical algebraic geometry, the arc space $X(D)$ of a scheme $X$ consists of maps from a formal disc $D = \text{Spec}(\mathbb{C}[[t]])$ to $X$. These are constructed as limits of truncated arc spaces $X(D_n)$ (maps from $\text{Spec}(\mathbb{C}[t]/t^{n+1})$).
• The Gap: Gaitsgory and Rozenblyum (in reference [4]) introduced derived arc spaces. They observed that for smooth schemes, the derived arc space is equivalent to the classical one. However, for singular schemes, the derived version can differ, carrying non-trivial higher homotopy sheaves (quasi-coherent sheaves representing the "derived thickening").
• The Question: The author investigates whether this derived thickening is a general phenomenon for singular schemes or if it vanishes for specific classes of singularities. Specifically, the author tests the hypothesis that derived arc spaces might relate to singularity invariants (like vanishing cycles) for singular hypersurfaces.
• The Goal: To determine the precise conditions under which the inclusion of classical arcs into derived arcs, $X(D)^{cl} \hookrightarrow X(D)$, is an equivalence.

2. Methodology

The author employs the language of pre-stacks and derived algebras (specifically non-positively graded commutative differential graded algebras, CDGAs) to construct explicit models.

• Explicit Co-fibrant Models:
  • The paper constructs explicit co-fibrant models for the algebras of functions on derived arc spaces.
  • If a derived scheme $X = \text{Spec}(A)$ is defined by a CDGA $A$ with generators $x_\lambda$ and relations $\partial(x_\lambda) = f_\lambda$, the algebra of functions on the arc space, $A(D)$, is generated by variables $x_\lambda^i$ (for $i \geq 0$) representing the Taylor coefficients of the arc.
  • The differential is defined via a generating function: $\partial(x_\lambda(t)) = f_\lambda(x_\mu(t))$, where $x_\lambda(t) = \sum x_\lambda^i t^i$.
• Grading and Filtration:
  • The author introduces a conformal weight grading (induced by the $\mathbb{G}_m$ action on the disc).
  • A filtration $F_{\leq n}$ is constructed based on the weight of the highest conformal degree variables.
  • This allows the analysis of the associated graded algebra, which is shown to be a symmetric algebra involving the cotangent complex.
• Spectral Sequences:
  • The author utilizes a convergent $E_1$ spectral sequence relating the homotopy groups of the symmetric algebra of the cotangent complex to the homotopy groups of the arc space algebra.
• Weak Smoothness Criterion:
  • A key intermediate concept is "weak smoothness." A scheme $X$ is weakly smooth if its cotangent complex $\mathbb{L}_X$ has no higher homotopy groups (i.e., $\mathbb{L}_X \cong \Omega_X^1[0]$).

3. Key Contributions and Theoretical Results

A. Characterization of Classicality (Theorem 5.2)

The paper establishes a necessary and sufficient condition for a derived arc space to be classical:

Theorem: For a classical scheme $X$, the derived arc space $X(D)$ is classical (equivalent to its classical truncation) if and only if $X$ is weakly smooth.

• Proof Logic:
  • If $X(D)$ is classical, the sub-complex of conformal weight 1 in the function algebra corresponds to the cotangent complex $\mathbb{L}_X$. If $X(D)$ has no higher homotopy, $\mathbb{L}_X$ must be concentrated in degree 0.
  • Conversely, if $X$ is weakly smooth, an inductive argument on the truncated arc spaces $X(D_n)$ using the filtration and spectral sequence shows that all higher homotopy groups vanish.

B. Application to Local Complete Intersections (Theorem 5.4)

The author applies the weak smoothness criterion to reduced local complete intersection (lci) schemes.

• Lemma 5.3: If $X$ is a reduced lci scheme, it is weakly smooth.
  • Reasoning: For a reduced complete intersection defined by equations $(f_1, \dots, f_c)$, the cotangent complex is computed via the Jacobian matrix. The conormal sequence is exact on the left for reduced complete intersections, implying $\pi_1(\mathbb{L}_X) = 0$. Since $\pi_{>1}$ vanishes trivially for lci schemes, the cotangent complex is concentrated in degree 0.
• Main Theorem (Theorem 2.1 / 5.4):

  If $X$ is a reduced local complete intersection scheme, the inclusion $X(D)^{cl} \hookrightarrow X(D)$ is an equivalence.

4. Results and Examples

• Non-Smooth Counter-examples: The paper confirms that for non-reduced or non-lci singularities, the derived arc space is distinct.
  • Example: For $X = \text{Spec}(\mathbb{C}[z]/z^2)$, the derived arc space has non-trivial higher homotopy (specifically $\pi_1 \neq 0$), generated by an element like $2x_1\zeta_0 - x_0\zeta_1$.
• The "Disappointing" Result: The author notes that the hope of relating the higher homotopy sheaves of derived arc spaces to interesting singularity invariants (like vanishing cycles cohomology) for singular hypersurfaces is unrealized in the case of reduced lci schemes. The derived structure collapses to the classical one for this broad and important class of singularities.
• Nilpotent Cone: The paper references the nilpotent cone (an lci scheme) as a case where the derived thickening is trivial, consistent with the main theorem.

5. Significance

• Clarification of Derived Geometry: The paper clarifies the scope of "derivedness" in arc spaces. It demonstrates that the complexity of derived arc spaces is not universal for all singularities but is strictly tied to the homological properties of the cotangent complex.
• Limitations of Derived Invariants: It serves as a cautionary result for researchers hoping to use derived arc spaces as a new tool to detect singularities via higher homotopy sheaves. For the large class of reduced lci schemes (which includes many standard examples in singularity theory), the derived arc space offers no new information beyond the classical arc space.
• Technical Framework: The construction of explicit co-fibrant models and the use of conformal weight filtrations provide a robust methodological toolkit for analyzing derived function algebras on infinite-dimensional spaces.

In summary, Bouaziz proves that derived arc spaces do not differ from classical arc spaces for any reduced local complete intersection scheme, effectively narrowing the class of schemes where derived arc spaces provide non-trivial higher homotopy information.