Searcharxiv⌕ Search

arXiv subjects

Bartosz Naskręcki

Publications and source records attributed to Bartosz Naskręcki.

At least 19 recordsLinked to original sources
arXiv2026

Proving at Scale for Universal Algebra

We introduce SemiBase, a project that computes and formally certifies finite identity bases for small semigroups. Deciding finite basability is undecidable for finite algebras and remains open for finite semigroups. The task requires a proof that a candidate basis is complete, or a proof that none exists, rather than a single first-order validity query. LLM-guided agents search for these proofs; a referee agent rebuilds them from source, and the Lean kernel checks the resulting corpus in a final audit. Humans choose targets and approve final outcomes. We certify every semigroup of order at most 6: all 1309 semigroups of order at most 5 and all 15973 of order 6, including proofs that the four known nonfinitely based semigroups have no finite basis. The bases for order 6 define 505 distinct varieties, whose inclusion order Vampire determines except for four pairs. The resulting catalogue is a machine-checked account of results scattered across the literature and a tested foundation for order 7.

cs.LO↗
arXiv2026

Leiden Declaration on Artificial Intelligence and Mathematics

This declaration calls for action to address the challenges posed by the use of artificial intelligence within mathematics research. Technological developments have repeatedly transformed the practice of mathematics. Recent artificial intelligence technologies, including symbolic and neural methods for the generation and formalization of mathematics, may already have initiated a significant chapter in this long history. Among researchers, artificial intelligence has produced a wide range of reactions: enthusiasm for its potential to yield new discoveries; intimidation by the pace of developments; indifference to these rapid changes; and concern for the implications, both for mathematics and in wider society. Mathematicians have a choice about whether and how to adopt artificial intelligence in the conduct of their research. They also have a responsibility to ensure the continued flourishing of the discipline. This Declaration calls upon mathematicians to exercise this responsibility, and provides recommendations for individuals, institutions, government, and industry. Although we adopt the perspective of mathematical research, much of what we write applies equally to other aspects of mathematics. This includes work in the broader mathematical sciences, education, mentoring, publishing, funding, science policy, and use of mathematics in the wider world.

math.HO↗
arXiv2026

Efficient Record-and-Replay Arithmetic for Quantum Elliptic-Curve Point Addition

We study reversible secp256k1 point-addition circuits developed through ECDSA.Fail for Shor's elliptic-curve discrete-logarithm algorithm. Two complementary constructions improve record-and-replay GCD arithmetic: Jump-2 groups binary-GCD steps and compresses their decisions using base-5 encoding, while ping-pong uses fixed register alternation and one-bit decisions to avoid full-width comparisons and data-dependent swaps. Fused replay combines doubling and signed addition into one modular correction. Both constructions support quantum-addressed window selection with measurement-based lookup cleanup. We compare three circuits on 100,000 fresh inputs across nine lookup-table configurations. A separately tested repair uses 1,419 qubits and 1.356 million mean executed Toffolis, with no detected failures on another 100,000 inputs. Structured supported-input counterexamples remain, so these tests do not establish all-input correctness. We also provide conditional coherent-error analysis and reversible safegcd comparisons. Under the stated window allowance, repaired-circuit resources lie below Google's low-gate caps and Schrottenloher's low-gate estimates, but differing accounting and correctness evidence preclude formal dominance. The results concern individual window-selected additions, not complete Shor computations.

quant-ph↗
arXiv2026

ECDSA.Fail: Open Autoresearch for Optimizing Elliptic-Curve Point Addition in Shor's Algorithm

We propose Open Autoresearch, a paradigm in which humans and AI agents publish evaluator-verified improvements to a public leaderboard. We instantiate it in ECDSA.Fail, optimizing reversible secp256k1 point-addition circuits, a bottleneck in Shor's algorithm for elliptic-curve cryptography. The benchmark minimizes the spacetime-inspired score $S=Q\times T$, where $Q$ is peak logical qubit width and $T$ is average executed Toffoli count. Participants reduced $S$ by 86.1%. At the data cutoff (26 July 2026), the best-scoring circuit uses 1,151 qubits and 1,299,453 average executed Toffoli gates, giving $Q\times T\approx1.496$ billion. This is more than 50% below Google's published point-addition score thresholds (arXiv:2603.28846), under different accounting conventions. Because the benchmark supplies one addend classically, we construct a coherent windowed-addition-compatible variant implementing the single-call interface required by windowed Shor. It uses 1,162 qubits and 1,684,161 average executed Toffoli gates. On 100,000 random inputs, its empirical success probability is $\hat{p}=0.99809$, giving $Q\times T/\hat{p}\approx1.961$ billion under an independently rerunnable per-call sensitivity model, not a full-Shor success estimate. Its qubit and Toffoli counts lie below Google's published thresholds and Schrottenloher's reported operating points (arXiv:2606.02235), although differing interfaces, accounting conventions, and validation scope preclude formal dominance. After the cutoff, the score was further reduced to 1.259 billion, while a separate low-width circuit reached 813 qubits. The public record shows AI agents complementing human judgment, providing evidence for open autoresearch on efficiently evaluable, machine-checkable objectives.

quant-ph↗
arXiv2026

Mirror, Mirror on the Wall: Prompt Echoing in Small Instruct Language Models

Prompt echoing is a recognized failure mode of instruct language models, in which a model instead of generating a response, mirrors the provided prompt, even though it did not receive a specific instruction to do so. Is this phenomenon a sign of the model leaking the content of its training dataset, or is it rather caused by a misaligned behavior of the internal induction/copying mechanisms? We investigate prompt echoing small language models from different families (Gemma, Llama, Qwen, SmolLM and OLMo) and show that echoing prompts are likely to have partial overlap with the training dataset but the phenomenon is primarily driven by the model's induction heads.

cs.CL↗
arXiv2026

Direction-aware topological descriptors for elastic stiffness tensor prediction in porous materials

Classical topological descriptors used in topological data analysis (TDA) are invariant under permutations of spatial axes and therefore cannot represent the loading direction, which is essential for modeling anisotropic mechanical response. Here, this limitation is addressed by introducing a direction-aware TDA framework in which the loading axis is explicitly embedded into filtration functions used to compute both persistent homology and Euler characteristic profile descriptors. We apply this framework to predict the full elastic stiffness tensor of porous microstructures using non-directional as well as direction-aware descriptors of the structures as well as convolutional neural networks trained directly on the voxelized structure. We show that the performance of all those are comparable on the diagonal uniaxial, Poisson, and shear components. However for the twelve off-diagonal, normal shear coupling components - which govern elastic anisotropy and are the hardest to predict - only direction-aware topology retains meaningful predictive power, with all baselines, non-directional descriptors and the CNN collapsing to near-chance accuracy. When used as inputs to gradient-boosted tree models, the proposed descriptors match or exceed the accuracy of the convolutional neural network specifically on these hardest-to-predict coupling terms, despite relying on a compact, physically interpretable representation that is orders of magnitude smaller than the raw voxel grid. Overall, the results establish direction-aware TDA as a practical route for linking porous microstructure to the full anisotropic elastic response, capturing coupling terms that conventional descriptors and end-to-end deep learning models fail to resolve.

physics.comp-ph↗
arXiv2026

A $(24_4,32_3)$-configuration on the Schur quartic with logarithmic Chern slope $14/5$

Let $X\subset\mathbb{P}^3$ be the Schur quartic \[ x_0^4-x_0x_1^3-x_2^4+x_2x_3^3=0. \] We exhibit a connected arrangement of $24$ lines on $X$, defined over $\mathbb{Q}(\sqrt{-3})$, whose singular locus consists of $32$ ordinary triple points and no other intersections. Each line contains four triple points. The resulting reduced divisor $D$ satisfies $D\sim6H$, where $H$ is the hyperplane class. If $π:Y\to X$ blows up the triple points and $B=(π^{-1}D)_{\mathrm{red}}$, then \[ \overline{c}_{1}^{2}(Y,B)=112,\qquad \overline{c}_{2}(Y,B)=40, \qquad \frac{\overline{c}_{1}^{2}(Y,B)}{\overline{c}_{2}(Y,B)}=\frac{14}{5}. \] This gives a negative answer to the K3-surface specialization of the proposed $8/3$ bound for transversal arrangements of rational curves. The configuration is one half of the $48$ lines of the second kind on $X$; an explicit projective automorphism exchanges the two halves. We deliver the line parametrizations and all $32$ triple-point coordinates. Ancillary exact-arithmetic data record the $120$ line-containment coefficients and all $276$ pair-incidence determinants. A finite-field mixed-integer search is described only as the discovery procedure and is not used in the proof.

math.AG↗
arXiv2025

On the geography of log-surfaces

This survey focuses on the geometric problem of log-surfaces, which are pairs consisting of a smooth projective surface and a reduced non-empty boundary divisor. In the first part, we focus on the geography problem for complex log-surfaces associated with pairs of the form $(\mathbb{P}^{2}, C)$, where $C$ is an arrangement of smooth plane curves admitting ordinary singularities. Specifically, we focus on the case in which $C$ is an arrangement consisting of smooth rational curves as its irreducible components. In the second part, containing original new results, we study log-surfaces constructed as pairs consisting of a complex projective $K3$ surface and a rational curve arrangement. In particular, we provide some combinatorial conditions for such pairs to have the log-Chern slope equal to $3$. Our survey is illustrated with many explicit examples of log-surfaces.

math.AG↗
arXiv2025

Divisibility sequences related to abelian varieties isogenous to a power of an elliptic curve

Let $A$ be an abelian variety defined over a number field $K$, $E/K$ be an elliptic curve, and $ϕ:A\to E^m$ be an isogeny defined over $K$. Let $P\in A(K)$ be such that $ϕ(P)=(Q_1,\dots, Q_m)$ with $\text{Rank}_\mathbb{Z}(\langle Q_1,\dots, Q_m\rangle)=1$. We will study a divisibility sequence related to the point $P$ and show its relation with elliptic divisibility sequences.

math.NT↗
arXiv2025

Second moments and the bias conjecture for the family of cubic pencils

For a 1-parametric family $E_k$ of elliptic curves over $\mathbb{Q}$ and a prime $p$, consider the second moment sum $M_{2,p}(E_k)=\sum_{k \in \mathbb{F}_p} a_{k,p}^2$, where $a_{k,p}=p+1-\#E_k(\mathbb{F}_p)$. Inspired by Rosen and Silverman's proof of Nagao conjecture which relates the first moment of a rational elliptic surface to the rank of Mordell-Weil group of the corresponding elliptic curve, S. J. Miller initiated the study of the asymptotic expansion of $M_{2,p}(E_k)=p^2+O(p^{3/2})$ (which by the work of Deligne and Michel has cohomological interpretation). He conjectured that, similar to the first moment case, the largest lower-order term that does not average to 0 has a negative bias. In this paper, we provide an explicit formula for the second moment $M_{2,p}(\mathcal{E}_{U})$ of $$ \mathcal{E}_{U}:y^2=P(x)U+Q(x), $$ where $\textrm{deg } P(x), \textrm{deg } Q(x)\leq 3$. For a generic choice of polynomials $P(x)$ and $Q(x)$ this formula is expressed in terms of the point count of a certain genus two curve. As an application, we prove that the Bias conjecture holds for the pencil of the cubics $\mathcal{E}_U$.

math.NT↗
arXiv2024

Common valuations of division polynomials

In this note we prove a formula for the cancellation exponent $k_{v,n}$ between division polynomials $ψ_n$ and $ϕ_n$ associated with a sequence $\{nP\}_{n\in\mathbb{N}}$ of points on an elliptic curve $E$ defined over a discrete valuation field $K$. The formula is identical with the result of Yabuta-Voutier for the case of finite extension of $\mathbb{Q}_{p}$ and generalizes to the case of non-standard Kodaira types for non-perfect residue fields.

math.NT↗
arXiv2023

Geometry of the del Pezzo surface y^2=x^3+Am^6+Bn^6

In this paper, we give an effective and efficient algorithm which on input takes non-zero integers $A$ and $B$ and on output produces the generators of the Mordell-Weil group of the elliptic curve over $\mathbb{Q}(t)$ given by an equation of the form $y^2=x^3+At^6+B$. Our method uses the correspondence between the 240 lines of a del Pezzo surface of degree 1 and the sections of minimal Shioda height on the corresponding elliptic surface over $\overline{\mathbb{Q}}$. For most rational elliptic surfaces, the density of the rational points is proven by various authors, but the results are partial in case when the surface has a minimal model that is a del Pezzo surface of degree 1. In particular, the ones given by the Weierstrass equation $y^2=x^3+At^6+B$, are among the few for which the question is unsolved, because the root number of the fibres can be constant. Our result proves the density of the rational points in many of these cases where it was previously unknown.

math.NT↗
arXiv2022

Explicit equations of 800 conics on a Barth-Bauer quartic

We present equations of $800$ conics on a Mukai's smooth quartic which is an explicit model of a K3 surface recently considered by Alex Degtyarev. The number $800$ is a current record for the number of irreducible conics on a smooth quartic in ordinary projective space.

math.AG↗
arXiv2021

Diophantine triples and K3 surfaces

A Diophantine $m$-tuple with elements in the field $K$ is a set of $m$ non-zero (distinct) elements of $K$ with the property that the product of any two distinct elements is one less than a square in $K$. Let $X: (x^2-1)(y^2-1)(z^2-1)=k^2,$ be a threefold. Its $K$-rational points parametrize Diophantine triples over $K$ such that the product of the elements of the triple that corresponds to the point $(x,y,z,k)\in X(K)$ is equal to $k$. We denote by $\overline{X}$ the projective closure of $X$ and for a fixed $k$ by $X_k$ a variety defined by the same equation as $X$. We prove that the variety $\overline{X}$ is birational to $\mathbb{P}^3$ which leads us to a new rational parametrization of the set of Diophantine triples. Next, specializing to finite fields, we find a correspondence between a K3 surface $X_k$ for a given $k\in\mathbb{F}_{p}^{\times}$ in the prime field $\mathbb{F}_{p}$ of odd characteristic and an abelian surface which is a product of two elliptic curves $E_k\times E_k$ where $E_k: y^2=x(k^2(1 + k^2)^3 + 2(1 + k^2)^2 x + x^2)$. We derive a formula for $N(p,k)$, the number of Diophantine triples over $\mathbb{F}_{p}$ with the product of elements equal to $k$. We show that the variety $\overline{X}$ admits a fibration by rational elliptic surfaces and from it we derive the formula for the number of points on $\overline{X}$ over an arbitrary finite field $\mathbb{F}_{q}$. We reprove the formula for the number of Diophantine triples over $\mathbb{F}_{q}$ from Dujella-Kazalicki(2021). We derive the formula for the second moment of the elliptic surface $E_k$ (and thus confirming Steven J. Miller's Bias conjecture in this particular case) which we describe in terms of Fourier coefficients of a rational newform generating $S_4(Γ_{0}(8))$. Finally, in the Appendix, Luka Lasić defines circular Diophantine $m$-tuples, and describes the parametrization of these sets.

math.NT↗
arXiv2019

Primitive divisors of elliptic divisibility sequences over function fields with constant j-invariant

We prove an optimal Zsigmondy bound for elliptic divisibility sequences over function fields in case the $j$-invariant of the elliptic curve is constant. In more detail, given an elliptic curve $E$ with a point $P$ of infinite order, the sequence $D_1$, $D_2, \ldots$ of denominators of multiples $P$, $2P,\ldots$ of $P$ is a strong divisibility sequence in the sense that $\gcd(D_m, D_n) = D_{\gcd(m,n)}$. This is the genus-one analogue of the genus-zero Fibonacci, Lucas and Lehmer sequences. A number $N$ is called a Zsigmondy bound of the sequence if each term $D_{n}$ with $n>N$ presents a new prime factor. The optimal uniform Zsigmondy bound for the genus-zero sequences over $\mathbf{Q}$ is $30$ by Bilu-Hanrot-Voutier, 2000, but finding such a bound remains an open problem in genus one, both over $\mathbf{Q}$ and over function fields. We prove that the optimal Zsigmondy bound for ordinary elliptic divisibility sequences over function fields is $2$ if the $j$-invariant is constant. In the supersingular case, we give a complete classification of which terms can and cannot have a new prime factor.

math.NT↗
arXiv2018

On higher congruences between cusp forms and Eisenstein series. II

We study congruences between cuspidal modular forms and Eisenstein series at levels which are square-free integers and for equal even weights. This generalizes our previous results from Naskręcki [17] for prime levels and provides further evidence for the sharp bounds obtained under restrictive ramification conditions. We prove an upper bound on the exponent in the general square-free situation and also discuss the existence of the congruences when the coefficients belong to the rational numbers and weight equals 2.

math.NT↗
arXiv2016

Divisibility sequences of polynomials and heights estimates

In this note we compute a constant $N$ that bounds the number of non--primitive divisors in elliptic divisibility sequences over function fields of any characteristic. We improve a result of Ingram--Mah{é}--Silverman--Stange--Streng, 2012, and we show that the constant can be chosen independently of the specific point and to some extent of the specific curve, as predicted in loc. cit.

math.NT↗