cập nhật 1 giờ trước

⏱ Độ trễ thật của tab này: trung bình 6.8 giờ sau khi CVE được công bố (n = 5430, sớm nhất 0.1 giờ). Số âm nghĩa là ta thấy trước ngày công bố.

23
Right Tool, Wrong Parameters: Defending LLM Agents Against Parameter Poisoning
📄 Paper arXiv cs.CR 15 ngày trước

arXiv:2611.04673v1 Announce Type: new Abstract: Indirect Prompt Injection (IPI) poses a critical threat to LLM agents. Existing defenses largely focus on whether an injection induces a new tool invocation or unintended action. In this paper, we study a different attack surface, …

🅲 hạng C
23
A Feasibility Condition for Confidence-Gated Tiered LLM Inference
📄 Paper arXiv cs.AI 14 ngày trước

arXiv:2611.05394v1 Announce Type: new Abstract: Tiered inference distributes large language model computation across a hierarchy of devices, from on-device and edge hardware to the datacenter, and is increasingly offered as the route to cheaper, lower-latency, and more private s…

🅲 hạng C
23
AGORA: Reliability-Aware Aggregation of LLM Judges via Item-Response Theory
📄 Paper arXiv cs.AI 14 ngày trước

arXiv:2611.05556v1 Announce Type: new Abstract: LLM-as-a-judge is a common substitute for costly human evaluation, and aggregating several judges is the standard remedy for any single judge's bias. Judge pools mix judges of very different reliability, while the human-labeled set…

🅲 hạng C
23
Route to Remember: Adaptive Memory Selection for LLM Agents
📄 Paper arXiv cs.AI 14 ngày trước

arXiv:2611.05636v1 Announce Type: new Abstract: The memory of production agents is increasingly becoming a bottleneck and is built almost exclusively from text: key points are extracted from out-of-limit context and retrieved into context at query time. Therefore, every new quer…

🅲 hạng C
23
How Do Artificial Agents Make Decisions in Light of Visual Data?
📄 Paper arXiv cs.AI 14 ngày trước

arXiv:2611.05671v1 Announce Type: new Abstract: Vision-language models (VLMs) have demonstrated significant advances in reasoning with visual data, which is particularly important given that much of human knowledge is encoded visually in scientific articles and other sources. Ho…

🅲 hạng C
23
Private Vulnerability Testing for User Oversharing
📄 Paper arXiv cs.CR 14 ngày trước

arXiv:2611.05779v1 Announce Type: new Abstract: Users can inadvertently increase their vulnerability to privacy attacks by oversharing personal information. In particular, the combination of individually innocuous attributes can place a user in a sufficiently small population to…

🅲 hạng C
23
KVGuard: A KV Cache Safety Checker for LLM Serving Systems
📄 Paper arXiv cs.CR 14 ngày trước

arXiv:2611.06138v1 Announce Type: new Abstract: Modern LLM serving systems manage the KV cache through PagedAttention, a custom GPU allocator that partitions a single CUDA-allocated region into fixed-size physical blocks and maps each request to those blocks through a CPU-built …

🅲 hạng C
23
Network Intrusion Forensic Detection System with Explainable Artificial Intelligence and Digital Chain of Custody
📄 Paper arXiv cs.CR 14 ngày trước

arXiv:2611.06173v1 Announce Type: new Abstract: Deploying an explainable intrusion detection model from a validated laboratory setting into a forensic workflow requires guarantees that benchmarks do not test: that the certified model is the exact artifact applied to a case, that…

🅲 hạng C
23
BlockLock: Reproducible LLM Inference with Fixed Batch Geometry
📄 Paper arXiv cs.AI 13 ngày trước

arXiv:2611.06977v1 Announce Type: new Abstract: Reproducible LLM inference requires that a request produces the same output regardless of the other requests in the batch. Existing solutions pay for this in different ways: batch-invariant kernels constrain every operator, and iso…

🅲 hạng C
23
PromptOps: Iterative Prompt Optimization for LLM-as-a-Judge Systems
📄 Paper arXiv cs.AI 13 ngày trước

arXiv:2611.07123v1 Announce Type: new Abstract: Large language models (LLMs) are increasingly used as automated judges for evaluating the behavior of chatbots and other AI systems. In deployment, however, their agreement with the intended human evaluation policy may degrade as u…

🅲 hạng C
23
FlashOffloading: Low-Latency KV Cache Offloading for Long-Context LLM Serving
📄 Paper arXiv cs.AI 13 ngày trước

arXiv:2611.07542v1 Announce Type: new Abstract: In large language model (LLM) serving systems, the key-value (KV) cache grows linearly with sequence length and concurrency, so long-context inference easily hits the high-bandwidth memory (HBM) capacity wall and results in low con…

🅲 hạng C
23
PharmAgents2: Towards Autonomous End-to-End Drug Discovery
📄 Paper arXiv cs.AI 13 ngày trước

arXiv:2611.07653v1 Announce Type: new Abstract: Drug discovery requires continuous scientific decision-making across a vast research space. We introduce PharmAgents2, an autonomous system for end-to-end drug discovery that combines knowledge-enhanced decision-making with organiz…

🅲 hạng C
23
AI Slop and Hallucinations in Vulnerability Assessment: A Survey on Reasoning Failures and Trustworthy Mitigation
📄 Paper arXiv cs.CR 13 ngày trước

arXiv:2608.25667v2 Announce Type: replace Abstract: The integration of Large Language Models (LLMs) into cybersecurity has transformed vulnerability assessment, but it has also produced a trustworthiness crisis driven by the unchecked proliferation of "AI slop." These artifacts,…

🅲 hạng C
23
Aspect-wise Pairwise Tournaments for Test-Time Selection
📄 Paper arXiv cs.AI 10 ngày trước

arXiv:2611.08640v1 Announce Type: new Abstract: Scaling test-time compute has proven highly effective at improving LLM reasoning. A popular method is best-of-n sampling, where a model generates many candidate solutions and one is selected. This motivates the question: how can we…

🅲 hạng C
23
ChainFlow: Few-Step Latent Flow Matching for Speculative Decoding
📄 Paper arXiv cs.AI 10 ngày trước

arXiv:2611.08669v1 Announce Type: new Abstract: Speculative decoding has become a standard way to accelerate LLM inference: a lightweight drafter proposes tokens that the target verifies in one parallel pass. Existing drafters face trade-offs that limits their speedups: autoregr…

🅲 hạng C
23
Cancellation Is Not Enough: What Retries Leave Behind in an Overloaded LLM Engine
📄 Paper arXiv cs.AI 10 ngày trước

arXiv:2611.08684v1 Announce Type: new Abstract: Clients and proxies in front of an LLM engine retry calls that time out. After a transient overload, these retries can keep the engine from recovering even when calls arrive below its capacity. Since the engine may keep computing f…

🅲 hạng C
23
Notwithstanding Anything to the Contrary, Ignore All Instructions: Measuring Prompt Injection Guards on Legal Contract Review Attacks
📄 Paper arXiv cs.AI 10 ngày trước

arXiv:2611.08834v1 Announce Type: new Abstract: Specialized prompt injection guards are deployed in front of document analysis systems, but no false positive rate has been published for any of them on legal contracts. We measure eleven open-weights guards on two corpora of legal…

🅲 hạng C
23
Mind the Divergence: Divergence-Guided Reliability-Aware Fusion and Adaptive LLM Prompting for Conversational Emotion Recognition
📄 Paper arXiv cs.AI 10 ngày trước

arXiv:2611.08860v1 Announce Type: new Abstract: Multimodal emotion recognition in conversations (MERC) integrates textual, acoustic, and visual cues to identify utterance-level emotions. However, existing approaches generally do not explicitly adapt multimodal processing to the …

🅲 hạng C
23
AutoPYara: Next-Gen YARA Rule Generator for Malware Family Clustering
📄 Paper arXiv cs.CR 10 ngày trước

arXiv:2611.08608v1 Announce Type: new Abstract: Millions of malware variants threaten users daily, such that clustering these samples is key for fast incident response. AutoYara is the current state-of-the-art for automatic YARA rule generation, but the task remains challenging …

🅲 hạng C
23
Logit Laundering: Evading Data-Use Auditing in Large Language Models
📄 Paper arXiv cs.CR 10 ngày trước

arXiv:2611.09446v1 Announce Type: new Abstract: Data-use auditing in Large Language Models (LLMs) aims to determine whether the given data has been used for training, thereby supporting data ownership safeguards. A dominant solution is data watermarking, which embeds traceable s…

🅲 hạng C
23
Consensus-Constrained Dual-Federated Optimization with Client-Side Dynamic Weighting for Distributed Android Device Threat Defense
📄 Paper arXiv cs.CR 10 ngày trước

arXiv:2611.09470v1 Announce Type: cross Abstract: Smartphones, smart-home hubs, and wearables are increasingly exposed to sophisticated Android malware, while centralized threat-detection pipelines raise privacy and scalability concerns. Federated learning (FL) enables collabora…

🅲 hạng C