[Paper Review] Exterior algebra methods for the Minimal Resolution Conjecture

David Eisenbud, Sorin Popescu+2 more|ArXiv.org|Nov 28, 2000

Advanced Numerical Analysis Techniques4 citations

TL;DR

This paper establishes that the Minimal Resolution Conjecture fails for general sets of γ points in Pr when r ≥ 6 and r ≠ 9, extending previous results by identifying approximately √2 − 1√r counterexamples per r using exterior algebra methods. The authors replace the complex degeneration argument in earlier proofs with a criterion based on 1-genericity and the BGG correspondence, enabling a simpler verification of irredundancy in linear complexes and recovering failures for linearly normal curves with d ≥ 3g − 2, g ≥ 4.

ABSTRACT

If r\geq 6, r eq 9, we show that the Minimal Resolution Conjecture fails for a general set of m points in P^r for almost 1/2\sqrt r values of m. This strengthens the result of Eisenbud and Popescu [1999], who found a unique such m for each r in the given range. Our proof begins like a variation of that of Eisenbud and Popescu, but uses exterior algebra methods as explained by Eisenbud and Schreyer [2000] to avoid the degeneration arguments that were the most difficult part of the Eisenbud-Popescu proof. Analogous techniques show that the Minimal Resolution Conjecture fails for linearly normal curves of degree d and genus g when d\geq 3g-2, g\geq 4, reproving results of Schreyer, Green, and Lazarsfeld.

Motivation & Objective

• To extend the known counterexamples to the Minimal Resolution Conjecture beyond the single γ value previously found by Eisenbud and Popescu.
• To replace the intricate degeneration argument in the original proof with a more direct criterion based on exterior algebra and 1-genericity.
• To show that the failure of the MRC for general sets of points in Pr is not isolated but occurs for a wide range of γ values when r ≥ 6, r ≠ 9.
• To recover and reprove the failure of the MRC for linearly normal curves of degree d ≥ 3g − 2 and genus g ≥ 4 using the same method.

Proposed method

• Utilizes the Gale transform duality between sets of points in Pr and Ps to relate the resolutions of their ideals.
• Applies the Bernstein-Gel’fand-Gel’fand (BGG) correspondence to study the irredundancy of linear free complexes.
• Employs exterior algebra techniques to derive a criterion for injectivity of a complex map φΓ based on the 1-genericity of multiplication maps in the canonical module.
• Uses Kreuzer’s result on 1-genericity of multiplication maps in the cone over a general set of points to verify the key condition without degeneration.
• Applies the theory to curves via a geometric 1-generic pairing between H0(ωC) and H0(L ⊗ ω−1C), constructing a complex whose irredundant quotient injects into the resolution.
• Verifies that the resulting Betti number inequalities violate the expected values under the MRC, proving failure.

Experimental results

Research questions

• RQ1 Does the Minimal Resolution Conjecture fail for more than one value of γ in Pr when r ≥ 6 and r ≠ 9?
• RQ2 Can the degeneration argument used in Eisenbud and Popescu’s proof be replaced with a more direct algebraic criterion?
• RQ3 What is the full range of γ for which the MRC fails for general sets of points in Pr when r ≥ 6, r ≠ 9?
• RQ4 Can the same method be applied to prove the failure of the MRC for linearly normal curves of degree d and genus g?
• RQ5 What is the precise condition under which the maximal irredundant quotient of a linear complex injects into a resolution, and how does this relate to Betti number violations?

Key findings

• The Minimal Resolution Conjecture fails for general sets of γ points in Pr when r ≥ 6, r ≠ 9, and r + 2 + √(r + 2) ≤ γ ≤ r + (3 + √(8r + 1))/2, yielding about (√2 − 1)√r counterexamples per r.
• The number of counterexamples per r is significantly larger than the single γ value previously found by Eisenbud and Popescu.
• The proof avoids the subtle degeneration argument by using a criterion based on exterior algebra and 1-genericity, simplifying the verification of irredundancy.
• The method successfully recovers the failure of the MRC for linearly normal curves of degree d ≥ 3g − 2 and genus g ≥ 4, including the case d = 3g − 2.
• For curves with d ≥ g² − g + 1, the failure is due to βd−2g+1,d−2g+2 > 1, which is detected via the Eagon-Northcott complex from a 1-generic pairing.
• The paper confirms Boij’s example of 21 points in P15 as a counterexample, which was not included in the original Eisenbud-Popescu list.