April 2026

arXiv:2604.18697v1 Announce Type: new Abstract: Indistinguishability properties such as differential privacy bounds or low empirically measured membership inference are widely treated as proxies to show a model is sufficiently protected against broader memorization risks. However, we show that indistinguishability properties are neither sufficient nor necessary for preventing data extraction in LLM APIs. We formalize a privacy-game separation between extraction and indistinguishability-based privacy, showing that indistinguishability and inextractability are incomparable: upper-bounding distinguishability does not upper-bound extractability. To address this […]

Ver mais

arXiv:2604.18667v1 Announce Type: new Abstract: We present linear-space data structures for several frequency queries on trees, namely: path mode, path least frequent element, and path $alpha$-minority queries. We present the first linear-space data structures, requiring $O(n sqrt{nw})$ preprocessing time, that can answer path mode and path least frequent element queries in $O(sqrt{n/w})$ time. This improves upon the best previously known bound of $O(loglog n sqrt{n/w})$ achieved by Durocher et al. in 2016. For the path $alpha$-minority problem, where […]

Ver mais

arXiv:2604.18665v1 Announce Type: new Abstract: This report presents an Audio-aware Referring Video Object Segmentation (Ref-VOS) pipeline tailored to the MEVIS_Audio setting, where the referring expression is provided in spoken form rather than as clean text. Compared with a standard Sa2VA-based Ref-VOS pipeline, the proposed system introduces two additional front-end stages: speech transcription and visual existence verification. Specifically, we first employ VibeVoice-ASR to convert long-form spoken input into a structured textual transcript. Since audio-derived queries are inherently noisy and […]

Ver mais

arXiv:2604.18663v1 Announce Type: new Abstract: Existing jamming attacks on Retrieval-Augmented Generation (RAG) systems typically induce explicit refusals or denial-of-service behaviors, which are conspicuous and easy to detect. In this work, we formalize a subtler availability threat, termed soft failure, which degrades system utility by inducing fluent and coherent yet non-informative responses rather than overt failures. We propose Deceptive Evolutionary Jamming Attack (DEJA), an automated black-box attack framework that generates adversarial documents to trigger such soft failures by exploiting […]

Ver mais

arXiv:2604.18661v1 Announce Type: new Abstract: We study a list-constrained extension of modular equation deletion over powers of two, called Coset-List Min-2-Lin$^{pm}$ over $mathbb{Z}/2^dmathbb{Z}$. Each variable is restricted to a dyadic coset $a+2^{ell}(mathbb{Z}/2^dmathbb{Z})$, each binary constraint is of the form $x_u=x_v$, $x_u=-x_v$, or $x_u=2x_v$, and the goal is to delete a minimum number of constraints so that the remaining system is satisfiable. This problem lies between the no-list case and the poorly understood fully conservative list setting. Our main […]

Ver mais

arXiv:2604.18660v1 Announce Type: new Abstract: Large Language Models (LLMs) are increasingly used in education, yet their default helpfulness often conflicts with pedagogical principles. Prior work evaluates pedagogical quality via answer leakage-the disclosure of complete solutions instead of scaffolding-but typically assumes well-intentioned learners, leaving tutor robustness under student misuse largely unexplored. In this paper, we study scenarios where students behave adversarially and aim to obtain the correct answer from the tutor. We evaluate a broad set of LLM-based tutor […]

Ver mais

arXiv:2604.18658v1 Announce Type: new Abstract: Existing AI agent safety benchmarks focus on generic criminal harm (cybercrime, harassment, weapon synthesis), leaving a systematic blind spot for a distinct and commercially consequential threat category: agents harming their own deployers. Real-world incidents illustrate the gap: Slack AI credential exfiltration (Aug 2024), Microsoft 365 Copilot calendar-injection leaks (Jan 2024), and a Meta agent unauthorized forum post exposing operational data (Mar 2026). We propose Owner-Harm, a formal threat model with eight categories of […]

Ver mais

arXiv:2604.18655v1 Announce Type: new Abstract: Deploying large language models (LLMs) on smartphones poses significant engineering challenges due to stringent constraints on memory, latency, and runtime flexibility. In this work, we present a hardware-aware framework for efficient on-device inference of a LLaMA-based multilingual foundation model supporting multiple use cases on Samsung Galaxy S24 and S25 devices with SM8650 and SM8750 Qualcomm chipsets respectively. Our approach integrates application-specific LoRAs as runtime inputs to a single frozen inference graph, enabling dynamic […]

Ver mais

arXiv:2604.18652v1 Announce Type: new Abstract: The transition of agentic AI from brittle prototypes to production systems is stalled by a pervasive crisis of craft. We suggest that the prevailing orchestration paradigm-delegating the system control loop to large language models and merely patching with heuristic guardrails-is the root cause of this fragility. Instead, we propose Arbiter-K, a Governance-First execution architecture that reconceptualizes the underlying model as a Probabilistic Processing Unit encapsulated by a deterministic, neuro-symbolic kernel. Arbiter-K implements a […]

Ver mais

arXiv:2604.18649v1 Announce Type: new Abstract: As generative AI faces intensifying legal challenges, the machine learning community has increasingly relied on post-hoc mitigation — especially machine unlearning and inference-time guardrails — to argue for compliance. This paper argues that such post-hoc mitigation methods cannot retroactively cure liability from unlawful acquisition and training, because compliance hinges on data lineage, not the outputs. Our argument has three parts. First, unauthorized copying/ingestion can be a legally complete completed act, and model weights […]

Ver mais