research note
This paper investigates how Android developers perceive platform efforts to reduce user tracking via device fingerprinting, a stealthy method that circumvents user controls and privacy protections
research note
The paper introduces the 'novelty bottleneck,' a conceptual framework that explains why human effort in AI-assisted tasks scales linearly with task size rather than sublinearly, despite advances in…
research note
This paper studies a question that most bot-detection work ignores: what happens when the bots are not static, but reinforcement-learning agents that adapt during the campaign? The authors use data…
research note
This paper addresses the challenge of training large language models (LLMs) for multi-step tool orchestration—executing sequences of API calls with interdependencies—where prior state-of-the-art su…
research note
This paper addresses the challenging problem of enabling native GUI agents—end-to-end vision-language models that operate directly on raw screenshots—to robustly solve modern interactive CAPTCHA ch…
research note
SkillProbe addresses critical security challenges in emerging Large Language Model (LLM) agent skill marketplaces, notably semantic-behavioral inconsistencies and inter-skill combinatorial risks
research note
This paper addresses the critical challenge of understanding and quantifying DRAM cell vulnerabilities that enable multiple security attacks, such as retention failures, Rowhammer, and Rowpress dis…
research note
This paper presents YouTube-Synch, a production-scale system designed for continuous, automated replication of YouTube video content from creator-authorized channels onto a blockchain-based decentr…
research note
This paper addresses the challenge of efficient and incentive-compatible routing in open, distributed multi-agent systems (MAS) built from large language model (LLM) inference agents
research note
This paper addresses a critical misalignment in authorization models for AI agents executing user natural-language (NL) tasks on the web
research note
This paper addresses a key limitation in AI-driven autoformalization of mathematics: while proof assistants ensure logical correctness of proofs via type checking, they do not guarantee that the fo…
research note
This paper is a Lean 4/Mathlib formalization of the classical planar isoperimetric inequality, proved via Hurwitz’s Fourier-analytic argument rather than via variational methods
