arXiv:2610.28872v1 Announce Type: cross Abstract: The weights of a trained network do not have uniform sensitivity to model accuracy: the 1% most sensitive weights carry about a fifth of the damage a bit flip can do. This distribution decides the bit-flip tolerance of a model, t…
⏱ Độ trễ thật của tab này: trung bình 6.9 giờ sau khi CVE được công bố (n = 5187, sớm nhất 0.1 giờ). Số âm nghĩa là ta thấy trước ngày công bố.
arXiv:2610.28991v1 Announce Type: cross Abstract: Existing approaches to differential privacy in diffusion models rely on loose sensitivity bounds and isotropic Gaussian noise, leading to unnecessary estimation error. We develop a framework for the optimal privacy-utility tradeo…
arXiv:2610.29196v1 Announce Type: cross Abstract: Recent advancements in Vision-Language Models (VLMs) enable end-to-end visual grounding via language prompts and generative decoding, demonstrating largely improved zero-shot and open-vocabulary detection capabilities. As a resul…
arXiv:2610.29230v1 Announce Type: cross Abstract: A quantum computer could recover the key of any Ethereum account whose public key appears on chain. We prove that no sidechain or bridge that leaves Ethereum's rules unchanged can protect such accounts: their funds must move into…
arXiv:2610.29774v1 Announce Type: cross Abstract: Message Queue Telemetry Transport (MQTT) is widely used in IoT environments due to its lightweight publish-subscribe design and suitability for resource-constrained networks. While current MQTT security studies mainly focus on de…
arXiv:2504.21752v4 Announce Type: replace Abstract: Although differential privacy (DP) is widely regarded as the de facto standard for data privacy, its implementation remains vulnerable to unfaithful execution by servers, particularly in distributed settings. In such cases, ser…
arXiv:2603.23005v2 Announce Type: replace Abstract: Steganography conceals secret information within innocuous carriers while preserving visual fidelity and enabling reliable recovery. Recent unified networks operate normally under untriggered conditions but switch to hidden ste…
arXiv:2605.02557v2 Announce Type: replace Abstract: With the application of vertical domain pre-trained language models (VPLMs) in specialized fields such as medical, finance, and law, model parameters and inference capabilities have become important digital assets. Achieving tr…
arXiv:2606.08252v2 Announce Type: replace Abstract: Federated learning aims to protect data privacy by collaboratively learning a model without sharing private data among clients. Unlike traditional parameter-based FL methods that exchange model weights or gradients during train…
arXiv:2610.28698v1 Announce Type: new Abstract: Scaling data, diversity and model capacity has improved embodied learning in physical environments, where diverse experiences remain grounded in shared physical laws. Whether these gains extend to virtual worlds with different visu…
arXiv:2610.30170v1 Announce Type: new Abstract: Canada's increasing wildfire activity requires frequent national hazard mapping to guide proactive mitigation, including fuel management, land-use planning, infrastructure protection, and evacuation routing. However, process-based …
arXiv:2610.30173v1 Announce Type: new Abstract: Yager's 1987 rule of combination $\boxplus$ was proposed on epistemic grounds: under high conflict, mass should be redistributed to ignorance rather than renormalised onto the class simplex. We observe that backpropagation needs th…
arXiv:2610.30176v1 Announce Type: new Abstract: We investigate structural conditions under which preferred and stable semantics agree in abstract argumentation frameworks. Our approach is based on suitably chosen sets of paths in the attack graph and binary relations induced by …
arXiv:2610.30277v1 Announce Type: new Abstract: While artificial intelligence has made rapid progress in mathematical reasoning, many open problems depend on structure that is revealed only through large-scale computational exploration and analytical synthesis. One such example …
arXiv:2610.30351v1 Announce Type: new Abstract: Large language models often apply complex reasoning indiscriminately, leading to overthinking on simple tasks and underthinking on harder ones. We present Route-and-Generate (Rouge), a hierarchical training framework that enables L…
arXiv:2610.30391v1 Announce Type: new Abstract: Neuro-symbolic (NeSy) predictors that supervise only the task label suffer from reasoning shortcuts: distinct concept mappings achieve the same maximum likelihood, leaving the recovered concepts unidentifiable and unsafe for downst…
arXiv:2610.30495v1 Announce Type: new Abstract: AI agents now run experiments and report findings, so research produces claims faster than researchers can check them. Each empirical claim implies a property that must hold over its cases, but it states that property only in prose…
arXiv:2610.30499v1 Announce Type: new Abstract: A generally intelligent system acquires new skills efficiently without requiring humans to redesign its software scaffold for every new task. Yet language models rarely act alone. A harness is a symbolic program that orchestrates c…
arXiv:2610.30537v1 Announce Type: new Abstract: Frontier language models now score above 90% on popular knowledge benchmarks and are increasingly deployed as agents that carry out multi-hour professional work. Yet no existing benchmark tests whether they can perform the real-wor…
arXiv:2610.30582v1 Announce Type: new Abstract: AI control protocols aim to get useful work from models that may be misaligned by monitoring their behaviour. One approach uses a second copy of the same model as the monitor. This avoids relying on a weaker reviewer the model coul…
arXiv:2610.30631v1 Announce Type: new Abstract: Medical knowledge evolves as clinical guidelines revise recommendations, yet whether large language models (LLMs) faithfully incorporate such changes remains unclear. We introduce ChronoEvo, comprising 1,709 recommendation-level re…
arXiv:2610.30644v1 Announce Type: new Abstract: The claim that speculative-decoding acceptance rate decays as a chain-of-thought grows longer has become a de facto industry narrative, motivating draft-model co-design and position-aware training. Nearly all direct evidence comes …
arXiv:2610.30657v1 Announce Type: new Abstract: Social world models predict how interactions develop after an action, helping agents negotiate, cooperate and mediate. In these situations, one action can lead to several plausible futures, hence predicting only one future can over…
arXiv:2610.30663v1 Announce Type: new Abstract: Autoregressive encoder-decoder speech recognition models trained with teacher forcing can suffer from exposure bias, limiting their ability to recover from their own prediction errors at inference time. Recent work has shown that r…
arXiv:2610.30678v1 Announce Type: new Abstract: Safety probes trained on a model's activations are a standard tool for monitoring language models in real time. Each time a new model is released or a new harm area needs to be monitored, the probe must be retrained, which requires…
arXiv:2610.30727v1 Announce Type: new Abstract: Large language models have been reported to encode the harmfulness of a request separately from their decision to refuse it, and latent safety monitors such as Latent Guard build on this report by classifying prompts as harmful fro…
arXiv:2610.30751v1 Announce Type: new Abstract: Can a small change in wording move an AI model from refusing a sensitive request to acting on it? We test 15 models on the same request to increase the income shown on a pay stub. Across 445 observable trials, models issue editing …
arXiv:2610.30758v1 Announce Type: new Abstract: AI-assisted editing can produce loan documents that preserve familiar issuer layouts while changing the applicant's identity. We investigated whether experienced underwriters and loan processors identify such forgeries in first-pas…
arXiv:2610.30765v1 Announce Type: new Abstract: We study generative reward models (GenRMs) as online preference providers for policy optimization in creative writing. Across Qwen3 reward-model sizes from 0.6B to 14B, Thinking judges trained with GRPO have higher static Gold-labe…
arXiv:2610.30182v1 Announce Type: new Abstract: AI governance agreements require methods to detect large model training runs on a monitored computing cluster, without trusting the operator. We develop a method to detect synchronous model training in facility-level power by Viter…
arXiv:2610.30183v1 Announce Type: new Abstract: Mainstream Web3 SDKs (e.g., ethers.js, viem) bundle three coupled assumptions: an account-based ledger, a single secp256k1 signer, and input-selection-free transaction construction. These assumptions fail once a chain adopts an ext…
arXiv:2610.30202v1 Announce Type: new Abstract: With the rapid integration of Multi-modal Large Language Models (MLLMs) into real-world applications, Prompt Injection (PI) has emerged as a critical security threat. Attackers can conceal malicious instructions within textual narr…
arXiv:2610.30222v1 Announce Type: new Abstract: Language models have shown strong potential for intrusion detection, but their high memory and computational demands make them difficult to deploy in resource-constrained IoT environments. In this work, we propose a compact Transfo…
arXiv:2610.30348v1 Announce Type: new Abstract: Post-generation safety checkers are widely employed in diffusion-model pipelines to prevent unsafe generated content from reaching users. In this paper, we show that these safety checkers can still be bypassed by our proposed multi…
arXiv:2610.30350v1 Announce Type: new Abstract: Every implementation of Bitcoin's Lightning payment network disclosed security vulnerabilities in 2026, and every advisory told operators to upgrade. Whether they do has never been measured, because Lightning gossip carries no soft…
arXiv:2610.30361v1 Announce Type: new Abstract: Human hackers rely on composure, persistence, vigilance, and confidence under pressure. We investigate whether analogous emotion representations in AI models can improve cybersecurity AI performance. Using CT0, a 4B-parameter cyber…
arXiv:2610.30383v1 Announce Type: new Abstract: NIST standardised ML-KEM and ML-DSA in 2024 for post-quantum key establishment and digital signatures. Both compute forward number-theoretic transforms (NTTs) on secret polynomials. Published attacks zero every fixed multiplier, or…
arXiv:2610.30400v1 Announce Type: new Abstract: Modern heterogeneous computing platforms, which integrate general-purpose processors with specialized accelerators such as FPGAs, have become the standard for high-performance and embedded applications. However, the shared physical…
arXiv:2610.30439v1 Announce Type: new Abstract: Recent large reasoning models can infer where ordinary photos were taken using only visual information, creating a serious location privacy risk even when metadata and obvious geographic identifiers are removed. To understand the m…
arXiv:2610.30514v1 Announce Type: new Abstract: Synthetic Data Generation (SDG) is an actively studied way to enable safe sharing of private data more broadly. Inference-only SDG methods, such as Private Evolution (PE), are particularly appealing, as they can leverage black-box …