Showing 1–7 of 7 results for author: Kural, M

The geometry of secondary terms in arithmetic statistics

Authors: Michael Kural

Abstract: In this thesis, we prove the existence of a secondary term for the count of cubic extensions of the function field $\mathbb{F}_q(t)$ of fixed absolute norm of discriminant. We show that the number of cubic extensions with absolute norm of discriminant equal to $q^{2N}$ is $c_1 q^{2N} - c_2^{i} q^{5N/3} + O_{\varepsilon}\left(q^{(3/2+\varepsilon)N}\right)$, where $c_1$ and $c_2^{i}$ are explicit co… ▽ More In this thesis, we prove the existence of a secondary term for the count of cubic extensions of the function field $\mathbb{F}_q(t)$ of fixed absolute norm of discriminant. We show that the number of cubic extensions with absolute norm of discriminant equal to $q^{2N}$ is $c_1 q^{2N} - c_2^{i} q^{5N/3} + O_{\varepsilon}\left(q^{(3/2+\varepsilon)N}\right)$, where $c_1$ and $c_2^{i}$ are explicit constants and $c_2^{i}$ only depends on $N\pmod{3}$. This builds on the work of Bhargava-Shankar-Tsimerman and Taniguchi-Thorne, who proved the existence of a secondary term for the count of cubic extensions of $\mathbb{Q}$ with bounded discriminant. Our approach uses a parametrization of Miranda and Casnati-Ekedahl, which can be seen as a geometric version of the classical parametrization by binary cubic forms used by Davenport-Heilbronn. This allows us to count and sieve for smooth curves embedded in Hirzebruch surfaces, in the same spirit as Zhao and Gunther. △ Less

Submitted 24 April, 2025; originally announced April 2025.

Comments: 65 pages. This was submitted (with minor formatting changes) to Harvard as the author's PhD thesis on April 3, 2025

MSC Class: 11R16 (Primary); 14J26; 11R58 (Secondary)

Quantifying Divergence for Human-AI Collaboration and Cognitive Trust

Authors: Müge Kural, Ali Gebeşçe, Tilek Chubakov, Gözde Gül Şahin

Abstract: Predicting the collaboration likelihood and measuring cognitive trust to AI systems is more important than ever. To do that, previous research mostly focus solely on the model features (e.g., accuracy, confidence) and ignore the human factor. To address that, we propose several decision-making similarity measures based on divergence metrics (e.g., KL, JSD) calculated over the labels acquired from… ▽ More Predicting the collaboration likelihood and measuring cognitive trust to AI systems is more important than ever. To do that, previous research mostly focus solely on the model features (e.g., accuracy, confidence) and ignore the human factor. To address that, we propose several decision-making similarity measures based on divergence metrics (e.g., KL, JSD) calculated over the labels acquired from humans and a wide range of models. We conduct a user study on a textual entailment task, where the users are provided with soft labels from various models and asked to pick the closest option to them. The users are then shown the similarities/differences to their most similar model and are surveyed for their likelihood of collaboration and cognitive trust to the selected system. Finally, we qualitatively and quantitatively analyze the relation between the proposed decision-making similarity measures and the survey results. We find that people tend to collaborate with their most similar models -- measured via JSD -- yet this collaboration does not necessarily imply a similar level of cognitive trust. We release all resources related to the user study (e.g., design, outputs), models, and metrics at our repo. △ Less

Submitted 18 January, 2024; v1 submitted 14 December, 2023; originally announced December 2023.

Transformers on Multilingual Clause-Level Morphology

Authors: Emre Can Acikgoz, Tilek Chubakov, Müge Kural, Gözde Gül Şahin, Deniz Yuret

Abstract: This paper describes our winning systems in MRL: The 1st Shared Task on Multilingual Clause-level Morphology (EMNLP 2022 Workshop) designed by KUIS AI NLP team. We present our work for all three parts of the shared task: inflection, reinflection, and analysis. We mainly explore transformers with two approaches: (i) training models from scratch in combination with data augmentation, and (ii) transf… ▽ More This paper describes our winning systems in MRL: The 1st Shared Task on Multilingual Clause-level Morphology (EMNLP 2022 Workshop) designed by KUIS AI NLP team. We present our work for all three parts of the shared task: inflection, reinflection, and analysis. We mainly explore transformers with two approaches: (i) training models from scratch in combination with data augmentation, and (ii) transfer learning with prefix-tuning at multilingual morphological tasks. Data augmentation significantly improves performance for most languages in the inflection and reinflection tasks. On the other hand, Prefix-tuning on a pre-trained mGPT model helps us to adapt analysis tasks in low-data and multilingual settings. While transformer architectures with data augmentation achieved the most promising results for inflection and reinflection tasks, prefix-tuning on mGPT received the highest results for the analysis task. Our systems received 1st place in all three tasks in MRL 2022. △ Less

Submitted 13 November, 2022; v1 submitted 3 November, 2022; originally announced November 2022.

Patterns of primes in the Sato-Tate conjecture

Authors: Nate Gillman, Michael Kural, Alexandru Pascadi, Junyao Peng, Ashwin Sah

Abstract: Fix a non-CM elliptic curve $E/\mathbb{Q}$, and let $a_E(p) = p + 1 - \#E(\mathbb{F}_p)$ denote the trace of Frobenius at $p$. The Sato-Tate conjecture gives the limiting distribution $μ_{ST}$ of $a_E(p)/(2\sqrt{p})$ within $[-1, 1]$. We establish bounded gaps for primes in the context of this distribution. More precisely, given an interval $I\subseteq [-1, 1]$, let $p_{I,n}$ denote the $n$th prim… ▽ More Fix a non-CM elliptic curve $E/\mathbb{Q}$, and let $a_E(p) = p + 1 - \#E(\mathbb{F}_p)$ denote the trace of Frobenius at $p$. The Sato-Tate conjecture gives the limiting distribution $μ_{ST}$ of $a_E(p)/(2\sqrt{p})$ within $[-1, 1]$. We establish bounded gaps for primes in the context of this distribution. More precisely, given an interval $I\subseteq [-1, 1]$, let $p_{I,n}$ denote the $n$th prime such that $a_E(p)/(2\sqrt{p})\in I$. We show $\liminf_{n\to\infty}(p_{I,n+m}-p_{I,n}) < \infty$ for all $m\ge 1$ for "most" intervals, and in particular, for all $I$ with $μ_{ST}(I)\ge 0.36$. Furthermore, we prove a common generalization of our bounded gap result with the Green-Tao theorem. To obtain these results, we demonstrate a Bombieri-Vinogradov type theorem for Sato-Tate primes. △ Less

Submitted 18 July, 2019; originally announced July 2019.

Comments: 26 pages

Journal ref: Res. Number Theory (2020) 6:9

Möbius formulas for densities of sets of prime ideals

Authors: Michael Kural, Vaughan McDonald, Ashwin Sah

Abstract: We generalize results of Alladi, Dawsey, and Sweeting and Woo for Chebotarev densities to general densities of sets of primes. We show that if $K$ is a number field and $S$ is any set of prime ideals with natural density $δ(S)$ within the primes, then \[ -\lim_{X \to \infty}\sum_{\substack{2 \le \operatorname{N}(\mathfrak{a})\le X\\ \mathfrak{a} \in D(K,S)}}\frac{μ(\mathfrak{a})}{\operatorname{N}(… ▽ More We generalize results of Alladi, Dawsey, and Sweeting and Woo for Chebotarev densities to general densities of sets of primes. We show that if $K$ is a number field and $S$ is any set of prime ideals with natural density $δ(S)$ within the primes, then \[ -\lim_{X \to \infty}\sum_{\substack{2 \le \operatorname{N}(\mathfrak{a})\le X\\ \mathfrak{a} \in D(K,S)}}\frac{μ(\mathfrak{a})}{\operatorname{N}(\mathfrak{a})} = δ(S), \] where $μ(\mathfrak{a})$ is the generalized Möbius function and $D(K,S)$ is the set of integral ideals $ \mathfrak{a} \subseteq \mathcal{O}_K$ with unique prime divisor of minimal norm lying in $S$. Our result can be applied to give formulas for densities of various sets of prime numbers, including those lying in a Sato-Tate interval of a fixed elliptic curve, and those in Beatty sequences such as $\lfloorπn\rfloor$. △ Less

Submitted 12 July, 2019; v1 submitted 5 July, 2019; originally announced July 2019.

Comments: 11 pages

Fuss-Schröder Paths and Rooted Plane Forests

Authors: Michael Kural

Abstract: We describe a bijection between $(k,k)$-Fuss-Schröder paths of type $λ$ and certain rooted plane forests with $n(k+1)+2$ vertices. This yields a recursion which allows us to analytically enumerate the number of large $(k,r)$-Fuss-Schröder paths of type $λ$, solving an open question posed by An, Jung, and Kim. Furthermore, we generalize the concept of $(k,r)$-Fuss-Schröder paths to $(k,S)$-Fuss-Sch… ▽ More We describe a bijection between $(k,k)$-Fuss-Schröder paths of type $λ$ and certain rooted plane forests with $n(k+1)+2$ vertices. This yields a recursion which allows us to analytically enumerate the number of large $(k,r)$-Fuss-Schröder paths of type $λ$, solving an open question posed by An, Jung, and Kim. Furthermore, we generalize the concept of $(k,r)$-Fuss-Schröder paths to $(k,S)$-Fuss-Schröder paths, in which $r$ can take any value in a given set $S$, and enumerate these paths as well. △ Less

Submitted 14 May, 2018; originally announced May 2018.

Dedekind Sums with Even Denominators

Authors: Michael Kural

Abstract: Let $S(a,b)$ denote the normalized Dedekind sum. We study the range of possible values for $S(a,b)=\frac{k}{q}$ with $\gcd(k,q)=1$. Girstmair proved local restrictions on $k$ depending on $q\pmod{12}$ and whether $q$ is a square and conjectured that these are the only restrictions possible. We verify the conjecture in the cases $q$ even, $q$ a square divisible by $3$ or $5$, and $2\le q\le 200$ (t… ▽ More Let $S(a,b)$ denote the normalized Dedekind sum. We study the range of possible values for $S(a,b)=\frac{k}{q}$ with $\gcd(k,q)=1$. Girstmair proved local restrictions on $k$ depending on $q\pmod{12}$ and whether $q$ is a square and conjectured that these are the only restrictions possible. We verify the conjecture in the cases $q$ even, $q$ a square divisible by $3$ or $5$, and $2\le q\le 200$ (the latter by computer), and provide progress towards a general approach. △ Less

Submitted 25 December, 2018; v1 submitted 1 January, 2018; originally announced January 2018.