arXiv subjects

Andy Zhang

Publications and source records attributed to Andy Zhang.

8 recordsLinked to original sources

arXiv2026

Let the Agent Steer: Closed-Loop Ranking Optimization via Influence Exchange

Recommendation ranking is fundamentally an influence allocation problem: a sorting formula distributes ranking influence among competing factors, and the business outcome depends on finding the optimal "exchange rates" among them. However, offline proxy metrics systematically misjudge how influence reallocation translates to online impact, with asymmetric bias across metrics that a single calibration factor cannot correct. We present Sortify, the first fully autonomous LLM-driven ranking optimization agent deployed in a large-scale production recommendation system. The agent reframes ranking optimization as continuous influence exchange, closing the full loop from diagnosis to parameter deployment without human intervention. It addresses structural problems through three mechanisms: (1) a dual-channel framework grounded in Savage's Subjective Expected Utility (SEU) that decouples offline-online transfer correction (Belief channel) from constraint penalty adjustment (Preference channel); (2) an LLM meta-controller operating on framework-level parameters rather than low-level search variables; (3) a persistent Memory DB with 7 relational tables for cross-round learning. Its core metric, Influence Share, provides a decomposable measure where all factor contributions sum to exactly 100%. Sortify has been deployed across two markets. In Country A, the agent pushed GMV from -3.6% to +9.2% within 7 rounds with peak orders reaching +12.5%. In Country B, a cold-start deployment achieved +4.15% GMV/UU and +3.58% Ads Revenue in a 7-day A/B test, leading to full production rollout.

cs.AI↗

arXiv2026

CTF for education

In this paper, we take a close look at how CTF can be used in cybersecurity education. We divide the CTF competitions into four different categories, which are attack-based CTFs, defense-based CTFs, jeopardy CTFs and gamified and wargames CTFs. We start our analysis by summarizing the main characteristics of different CTF types. We then compare them with each other in both learning objectives and other aspects like accessibility. We conclude that combining all four CTF formats can help participants build one's cybersecurity knowledge. By doing that, we hope that our findings will provide some useful insights for future CTF educators.

cs.CR↗

arXiv2025

The EBLM Project XVIII. 3D Obliquities of Five Low-Mass Eclipsing Binaries

The formation of tight stellar binaries remains an unsolved problem. There is too much angular momentum in a collapsing and fragmenting protostellar cloud to form a stellar binary in situ with a separation less than an AU, yet thousands of these short-period binaries have been discovered. One indication of a binary's formation is the angle between the stellar spin and orbital axes --- its obliquity. The classical method for determining projected stellar obliquities is the Rossiter-McLaughlin effect. This has been applied to 132 hot Jupiters, but only a handful of stellar binaries. Of the binary systems with measured projected obliquities, even fewer have measured 3D obliquities. In this paper, we add five more short-period binary 3D obliquity measurements to the sample previously consisting of one system. We present Rossiter-McLaughlin measurements for EBLM J0239-20, EBLM J0941-31, EBLM J1037-25, EBLM J1141-37, and EBLM J2025-45. These systems consist of an M-dwarf eclipsing an F/G type primary. We combine CORALIE and HARPS spectroscopy with TESS photometry of primary and secondary eclipses. We show that even though the sky-projected obliquities seem to be aligned, there is modest but non-zero spin-orbit misalignment ($\psi$ between 5 and 20$^{\circ}$). Our primary stars straddle the Kraft break at $\sim 6250K$. We derive the M-dwarf masses and radii to precisions better than 3\%. With the exception of EBLM J0941-31, each system has an inflated radius, exceeding stellar model predictions by more than 5$\sigma$.

astro-ph.SR↗

arXiv2025

SeedAIchemy: LLM-Driven Seed Corpus Generation for Fuzzing

We introduce SeedAIchemy, an automated LLM-driven corpus generation tool that makes it easier for developers to implement fuzzing effectively. SeedAIchemy consists of five modules which implement different approaches at collecting publicly available files from the internet. Four of the five modules use large language model (LLM) workflows to construct search terms designed to maximize corpus quality. Corpora generated by SeedAIchemy perform significantly better than a naive corpus and similarly to a manually-curated corpus on a diverse range of target programs and libraries.

cs.CR↗

arXiv2025

The EBLM project XVI. Moderate spin-orbit misalignment of the low mass eclipsing binary EBLM J0021-16

Thousands of tight ($<1$ AU) main sequence binaries have been discovered, but it is uncertain how they formed. There is likely too much angular momentum in a collapsing, fragmenting protostellar cloud to form such binaries in situ, suggesting some post processing. One probe of a binary's dynamical history is the angle between the stellar spin and orbital axes -- its obliquity. The classical method for determining stellar obliquity is the Rossiter-McLaughlin effect. It has been applied to over 100 hot Jupiters, but less than a dozen stellar binaries. In this paper, we present the Rossiter-McLaughlin measurement of EBLM J0021-16, a $0.19M_\odot$ M-dwarf eclipsing a $1.05M_\odot$ G-dwarf on a 5.97 day, almost-circular orbit. We combine CORALIE spectroscopy with TESS photometry and a measured primary star rotation period of 7.04 days, according to star spot modulation. We show that the orbital axis is misaligned with the primary star's spin axis, with a true 3D obliquity of $\psi=28.9\pm2.1^{\circ}$. EBLM J0021-16, being neither spin-orbit aligned nor synchronized, yet with an almost circular orbit, is a curious case for tidal evolution in tight binaries. It becomes one of a handful of eclipsing binaries with true obliquity measurements. Finally, we derive the M-dwarf's mass and radius to a fractional precision better than 1\%. The radius of the M-dwarf is inflated by 6\% ($7.4\sigma$) with respect to stellar models, consistent with many other M-dwarfs in the literature.

astro-ph.SR↗

arXiv2025

Establishing Best Practices for Building Rigorous Agentic Benchmarks

Benchmarks are essential for quantitatively tracking progress in AI. As AI agents become increasingly capable, researchers and practitioners have introduced agentic benchmarks to evaluate agents on complex, real-world tasks. These benchmarks typically measure agent capabilities by evaluating task outcomes via specific reward designs. However, we show that many agentic benchmarks have issues in task setup or reward design. For example, SWE-bench Verified uses insufficient test cases, while TAU-bench counts empty responses as successful. Such issues can lead to under- or overestimation of agents' performance by up to 100% in relative terms. To make agentic evaluation rigorous, we introduce the Agentic Benchmark Checklist (ABC), a set of guidelines that we synthesized from our benchmark-building experience, a survey of best practices, and previously reported issues. When applied to CVE-Bench, a benchmark with a particularly complex evaluation design, ABC reduces the performance overestimation by 33%.

cs.AI↗

arXiv2025

Frontier AI's Impact on the Cybersecurity Landscape

The impact of frontier AI (i.e., AI agents and foundation models) in cybersecurity is rapidly increasing. In this paper, we comprehensively analyze this trend through multiple aspects: quantitative benchmarks, qualitative literature review, empirical evaluation, and expert survey. Our analyses consistently show that AI's capabilities and applications in attacks have exceeded those on the defensive side. Our empirical evaluation of widely used agent systems on cybersecurity benchmarks highlights that current AI agents struggle with flexible workflow planning and using domain-specific tools for complex security analysis -- capabilities particularly critical for defensive applications. Our expert survey of AI and security researchers and practitioners indicates a prevailing view that AI will continue to benefit attackers over defenders, though the gap is expected to narrow over time. These results show the urgent need to evaluate and mitigate frontier AI's risks, steering it towards benefiting cyber defenses. Responding to this need, we provide concrete calls to action regarding: the construction of new cybersecurity benchmarks, the development of AI agents for defense, the design of provably secure AI agents, the improvement of pre-deployment security testing and transparency, and the strengthening of user-oriented education and defenses. Our paper summary and blog are available at https://rdi.berkeley.edu/frontier-ai-impact-on-cybersecurity/.

cs.CR↗

arXiv2024

Moment-based metrics for molecules computable from cryo-EM images

Single particle cryogenic electron microscopy (cryo-EM) is an imaging technique capable of recovering the high-resolution 3-D structure of biological macromolecules from many noisy and randomly oriented projection images. One notable approach to 3-D reconstruction, known as Kam's method, relies on the moments of the 2-D images. Inspired by Kam's method, we introduce a rotationally invariant metric between two molecular structures, which does not require 3-D alignment. Further, we introduce a metric between a stack of projection images and a molecular structure, which is invariant to rotations and reflections and does not require performing 3-D reconstruction. Additionally, the latter metric does not assume a uniform distribution of viewing angles. We demonstrate uses of the new metrics on synthetic and experimental datasets, highlighting their ability to measure structural similarity.

q-bio.BM↗