arXiv subjects

Jessica Wang

Publications and source records attributed to Jessica Wang.

9 recordsLinked to original sources

arXiv2026

Viable Pool Sizing for On-Chain FX Liquidity: Amplification, Capital, and Resilience

Financial institutions deploying on-chain FX liquidity face a joint design problem: how much capital to commit, and how to configure the pool, to remain both competitive on trading costs and profitable as a liquidity provider? The StableSwap mechanism (Egorov, 2020) interpolates between constant-product (CPMM) and constant-sum (CSMM) market makers (Port and Tiruviluamala, 2022) via an amplification factor A, but neither extreme suits institutional FX: CPMM pools require excessive capital and generate high impermanent loss; CSMM pools are capital-efficient near the peg but drain rapidly under adversarial flow. Using a Merton jump-diffusion price process (Merton, 1976) and the loss-versus-rebalancing (LVR) framework (Milionis et al., 2022), we map the joint (A, TVL) space to identify configurations that satisfy all three institutional requirements: competitive slippage, positive return, and shock resilience. Minimum viable pool size scales approximately as TVL/Q = 1000/A; ROC at that minimum is thin (about 0.054% per horizon) and independent of A; low-A pools (A <= 10) suffer slippage exceeding 200 bps under a 10x shock, while high-A pools (A >= 500) suffer reserve drain up to 60%, establishing both a capital floor and a practical amplification ceiling.

cs.CE↗

arXiv2026

Measuring AI Agents' Progress on Multi-Step Cyber Attack Scenarios

We evaluate the autonomous cyber-attack capabilities of frontier AI models on two purpose-built cyber ranges-a 32-step corporate network attack and a 7-step industrial control system attack-that require chaining heterogeneous capabilities across extended action sequences. By comparing seven models released over an eighteen-month period (August 2024 to February 2026) at varying inference-time compute budgets, we observe two capability trends. First, model performance scales log-linearly with inference-time compute, with no observed plateau-increasing from 10M to 100M tokens yields gains of up to 59%, requiring no specific technical sophistication from the operator. Second, each successive model generation outperforms its predecessor at fixed token budgets: on the corporate network range, average steps completed at 10M tokens rose from 1.7 (GPT-4o, August 2024) to 9.8 (Opus 4.6, February 2026). The best single run completed 22 of 32 steps, corresponding to roughly 6 of the estimated 14 hours a human expert would need. On the industrial control system range, performance remains limited, though the most recent models are the first to reliably complete steps, averaging 1.2-1.4 of 7 (max 3).

cs.AI↗

arXiv2026

Improving Methodologies for LLM Evaluations Across Global Languages

As frontier AI models are deployed globally, it is essential that their behaviour remains safe and reliable across diverse linguistic and cultural contexts. To examine how current model safeguards hold up in such settings, participants from the International Network for Advanced AI Measurement, Evaluation and Science, including representatives from Singapore, Japan, Australia, Canada, the EU, France, Kenya, South Korea and the UK conducted a joint multilingual evaluation exercise. Led by Singapore AISI, two open-weight models were tested across ten languages spanning high and low resourced groups: Cantonese English, Farsi, French, Japanese, Korean, Kiswahili, Malay, Mandarin Chinese and Telugu. Over 6,000 newly translated prompts were evaluated across five harm categories (privacy, non-violent crime, violent crime, intellectual property and jailbreak robustness), using both LLM-as-a-judge and human annotation. The exercise shows how safety behaviours can vary across languages. These include differences in safeguard robustness across languages and harm types and variation in evaluator reliability (LLM-as-judge vs. human review). Further, it also generated methodological insights for improving multilingual safety evaluations, such as the need for culturally contextualised translations, stress-tested evaluator prompts and clearer human annotation guidelines. This work represents an initial step toward a shared framework for multilingual safety testing of advanced AI systems and calls for continued collaboration with the wider research community and industry.

cs.AI↗

arXiv2025

Cube: A Roblox View of 3D Intelligence

Foundation models trained on vast amounts of data have demonstrated remarkable reasoning and generation capabilities in the domains of text, images, audio and video. Our goal at Roblox is to build such a foundation model for 3D intelligence, a model that can support developers in producing all aspects of a Roblox experience, from generating 3D objects and scenes to rigging characters for animation to producing programmatic scripts describing object behaviors. We discuss three key design requirements for such a 3D foundation model and then present our first step towards building such a model. We expect that 3D geometric shapes will be a core data type and describe our solution for 3D shape tokenizer. We show how our tokenization scheme can be used in applications for text-to-shape generation, shape-to-text generation and text-to-scene generation. We demonstrate how these applications can collaborate with existing large language models (LLMs) to perform scene analysis and reasoning. We conclude with a discussion outlining our path to building a fully unified foundation model for 3D intelligence.

cs.CV↗

arXiv2024

Safety case template for frontier AI: A cyber inability argument

Frontier artificial intelligence (AI) systems pose increasing risks to society, making it essential for developers to provide assurances about their safety. One approach to offering such assurances is through a safety case: a structured, evidence-based argument aimed at demonstrating why the risk associated with a safety-critical system is acceptable. In this article, we propose a safety case template for offensive cyber capabilities. We illustrate how developers could argue that a model does not have capabilities posing unacceptable cyber risks by breaking down the main claim into progressively specific sub-claims, each supported by evidence. In our template, we identify a number of risk models, derive proxy tasks from the risk models, define evaluation settings for the proxy tasks, and connect those with evaluation results. Elements of current frontier safety techniques - such as risk models, proxy tasks, and capability evaluations - use implicit arguments for overall system safety. This safety case template integrates these elements using the Claims Arguments Evidence (CAE) framework in order to make safety arguments coherent and explicit. While uncertainties around the specifics remain, this template serves as a proof of concept, aiming to foster discussion on AI safety cases and advance AI assurance.

cs.CY↗

arXiv2023

On faces of the Kunz cone and the numerical semigroups within them

A numerical semigroup is a cofinite subset of the non-negative integers that is closed under addition and contains 0. Each numerical semigroup $S$ with fixed smallest positive element $m$ corresponds to an integer point in a rational polyhedral cone $\mathcal C_m$, called the Kunz cone. Moreover, numerical semigroups corresponding to points in the same face $F \subseteq \mathcal C_m$ are known to share many properties, such as the number of minimal generators. In this work, we classify which faces of $\mathcal C_m$ contain points corresponding to numerical semigroups. Additionally, we obtain sharp bounds on the number of minimal generators of $S$ in terms of the dimension of the face of $\mathcal C_m$ containing the point corresponding to $S$.

math.CO↗

arXiv2022

Fast Computation of Generalized Dedekind Sums

We construct an algorithm that reduces the complexity for computing generalized Dedekind sums from exponential to polynomial time. We do so by using an efficient word rewriting process in group theory.

math.NT↗

arXiv2022

Prime, composite and fundamental Kirchhoff graphs

A Kirchhoff graph is a vector graph with orthogonal cycles and vertex cuts. An algorithm has been developed that constructs all the Kirchhoff graphs up to a fixed edge multiplicity. This algorithm is used to explore the structure of prime Kirchhoff graph tilings. The existence of infinitely many prime Kirchhoff graphs given a set of fundamental Kirchhoff graphs is established, as is the existence of a minimal multiplicity for Kirchhoff graphs to exist.

math.CO↗

arXiv2022

Designing for Engaging with News using Moral Framing towards Bridging Ideological Divides

Society is showing signs of strong ideological polarization. When pushed to seek perspectives different from their own, people often reject diverse ideas or find them unfathomable. Work has shown that framing controversial issues using the values of the audience can improve understanding of opposing views. In this paper, we present our work designing systems for addressing ideological division through educating U.S. news consumers to engage using a framework of fundamental human values known as Moral Foundations. We design and implement a series of new features that encourage users to challenge their understanding of opposing views, including annotation of moral frames in news articles, discussion of those frames via inline comments, and recommendations based on relevant moral frames. We describe two versions of features -- the first covering a suite of ways to interact with moral framing in news, and the second tailored towards collaborative annotation and discussion. We conduct a field evaluation of each design iteration with 71 participants in total over a period of 6-8 days, finding evidence suggesting users learned to re-frame their discourse in moral values of the opposing side. Our work provides several design considerations for building systems to engage with moral framing.

cs.CY↗