arXiv:2611.17534v1 Announce Type: new Abstract: Large language model (LLM) agents are increasingly used for investigative tasks over incomplete evidence, such as cybersecurity alert triage, but reinforcement learning (RL) for such tasks is difficult: intermediate steps have no v…
⏱ Độ trễ thật của tab này: trung bình 7.3 giờ sau khi CVE được công bố (n = 5337, sớm nhất 0.1 giờ). Số âm nghĩa là ta thấy trước ngày công bố.
arXiv:2508.00910v3 Announce Type: replace Abstract: Large Language Models (LLMs) have achieved remarkable success in software engineering tasks when trained with executable runtime environments, particularly in resolving GitHub issues. However, such runtime environments are ofte…
Google on Monday revealed that adversarial use of artificial intelligence (AI) by threat actors has gone beyond manual prompting to deploying autonomous multi-agent offensive frameworks like ARTEX, CAIRN, CyberStrikeAI, and PentAGI to execute multi-stage intrusions at greater ope…
0
Plugin WP Ultimate CSV Importer có lỗ hổng SQL injection khi nhập dữ liệu LifterLMS, không escape values, cho phép người dùng có quyền admin thực thi câu lệnh SQL arbitrary.
🎯 Kẻ tấn công làm gì (từng bước)
- Dò tham số/ô nhập được đưa vào câu lệnh SQL
- Chèn câu SQL để thử (lỗi cú pháp, OR 1=1, UNION SELECT, sleep)
- Dùng UNION/stacked query để đọc bảng users, token, đôi khi ghi file để chiếm máy
🔍 Dấu hiệu bị tấn công (để kiểm tra log)
- Log web có chuỗi đặc trưng: UNION SELECT, ' OR 1=1, sleep(, information_schema
- Tăng đột biến lỗi CSDL 500 ở cùng một tham số
- Một truy vấn kéo về số dòng bất thường (dump dữ liệu)
🛡 Phòng ngay (làm được hôm nay)
- Vá plugin/CMS có lỗi (với WordPress: xoá plugin không dùng, cập nhật plugin bị ảnh hưởng)
- Bật WAF rule chống SQLi (OWASP CRS) ở chế độ chặn
- Tài khoản CSDL của ứng dụng chỉ có quyền cần thiết (không DROP/GRANT)
- Bật log chậm/truy vấn bất thường để phát hiện dump
arXiv:2611.17856v1 Announce Type: new Abstract: Tools for interpreting the internal representations of Large Language Models (LLMs) increasingly rely on LLMs themselves to both generate and evaluate explanations, creating a self-referential automated interpretation process with …
Large language models (LLMs) are increasingly applied to finance, creating a need for benchmarks that evaluate their capabilities across financial systems with different regulations and practices. However, existing financial benchmarks provide limited coverage of country-specific…
Release: datasette-agent 0.6a0 Accumulated small fixes for Datasette Agent, my agent plugin for Datasette that lets you ask questions about your data. Most notably, it now adds an agent prompt directly to the Datasette homepage, with support for hash-mentions of tables and querie…
It is common to train Large Language Models (LLMs) for agentic tasks with Reinforcement Learning (RL) from scalar rewards. However, this throws away the fine-grained information about the environment that the agent experienced during the trial (e.g., error logs). To utilize this …
arXiv:2611.13696v1 Announce Type: new Abstract: Multi-agent systems provide a convenient framework for the massive deployment of agents based on Large Language Models (LLMs). It provides a sound framework where agents can interact with existing software to perform actions in an …
Efficient batch execution of agent tasks is critical to LLM training and development workflows, including benchmark evaluation, synthetic-data generation, and RL rollouts. Since these tasks alternate between LLM inference and environment interaction, we present coordinated optimi…
arXiv:2611.10411v1 Announce Type: new Abstract: LLM-based agents extend LLMs with tool calls, memory, planning, and retrieval to act in external environments, but they are vulnerable to attacks at different workflow steps. Existing defenses, including prompt engineering and post…
arXiv:2611.10756v1 Announce Type: new Abstract: Large Language Models (LLMs) are increasingly deployed as shared multi-tenant services, relying on advanced optimizations like KV cache and prefix sharing to improve throughput and reduce latency. However, these optimizations intro…
arXiv:2611.11104v1 Announce Type: new Abstract: As a core protocol for integrating large language models (LLMs) with external data sources and tools, the Model Context Protocol (MCP) has quickly become a key interface in emerging AI agent systems. As the MCP ecosystem expands, r…
Release: llm 0.37.3 A bug fix release for LLM , mainly to cover a problem with the most recent Homebrew package of the software. Homebrew now ships the latest SQLite, 3.54.1 , which includes a fix that broke LLM's logging code . Short version: in SQLite 3.54.0 the following SQL: …
Datasette-llm 0.1a11 giới thiệu plugin trung gian cho Datasette, hỗ trợ cấu hình LLM qua giao diện OpenAI-compatible, cho phép dùng mô hình local như LM Studio, Ollama, llama.cpp.
OverServe: kỹ thuật resource amplification attack nhắm vào SaaS agents — tenant xấu dùng prompt khai thác quyết định 'khi nào agent tự dừng' để kéo dài execution và đốt tài nguyên model/tool của provider (đặc biệt model subscription-based). Đáng chú ý với ai xây/thẩm định AI agent trong SaaS vì dễ bị lạm dụng chi phí.
🎯 Kẻ tấn công làm gì (từng bước)
- Tìm đúng đường dẫn/API có xử lý dữ liệu người dùng gửi lên
- Gửi payload đã chế để máy chủ thực thi lệnh/mã
- Dùng quyền của tiến trình web để đọc file, cài backdoor, mở kênh ra ngoài
🔍 Dấu hiệu bị tấn công (để kiểm tra log)
- Tiến trình web (php-fpm/nginx/java) sinh tiến trình con bất thường: sh, bash, curl, wget
- Kết nối ra ngoài từ máy chủ tới IP lạ đúng lúc có request lạ
- File mới xuất hiện trong thư mục tải lên/tạm
🛡 Phòng ngay (làm được hôm nay)
- Vá ngay bản có lỗi (ưu tiên số 1)
- Chặn tạm đường dẫn/API bị lỗi ở WAF/reverse proxy nếu chưa vá được
- Chạy tiến trình web bằng user quyền thấp, tắt các hàm shell (disable_functions) nếu dùng PHP
- Bật giám sát tiến trình con của web server để phát hiện lệnh lạ
arXiv:2611.17610v1 Announce Type: new Abstract: Evaluations of multi-agent language models often compare communicating architectures against a single greedy pass. This practice confounds coordination with sample multiplicity and inference compute. We synthesize nine primary empi…
arXiv:2611.17395v1 Announce Type: new Abstract: Recent advances in weight-space programming (WSP) have demonstrated that large language models (LLMs) can directly inspect and modify neural network weights without conventional gradient-based training. While this capability opens …
Since 2021, Project Zero has followed a “90 + 30” vulnerability disclosure policy, allowing vendors up to 90 days to remediate a vulnerability plus an extra 30 days “soak time” for patch rollout. Our goal was to encourage faster patching and give vendors time to deliver patches t…
arXiv:2611.15460v1 Announce Type: new Abstract: Underspecified instructions leave LLM agents to decide what to do. In public services, staff write these instructions, but the agent's decisions affect residents, patients, and students. We build four expert-validated public-servic…
arXiv:2611.12966v1 Announce Type: new Abstract: Large language models (LLMs) can generate SQL and SQL--Python workflows for investment research, but ensuring point-in-time (PIT) validity remains challenging. Agents must enforce disclosure cutoffs and select the appropriate histo…
arXiv:2605.14786v2 Announce Type: replace Abstract: As LLM-based agents increasingly browse the web on users' behalf, a natural question arises: can websites passively identify which underlying model powers an agent? Doing so would represent a significant security risk, enabling…
arXiv:2611.17818v1 Announce Type: new Abstract: Jailbreak attacks can circumvent the safety mechanisms of Large Language Models (LLMs) and lead to the generation of harmful or policy-violating content. We first examine eight jailbreak methods against an LLM-based filter and obse…
Large language models (LLMs) are increasingly used by reviewers to highlight weaknesses during the reviewing process, and their outputs often shape the reviews that humans submit. Prior work studies how LLM reviews align with human ones, yet we know little about which parts of a …
arXiv:2611.11399v1 Announce Type: new Abstract: Large Language Models (LLMs) present significant challenges to source tracing of generated code, as LLM rewriting can substantially alter the syntactic and structural characteristics of watermarked code. However, existing watermark…
We break down the Citrus Ransomware group's attacks: the attackers' TTPs and tooling, including VLESS traffic tunneling and LLM use, plus their links to Toy Ghouls.
arXiv:2508.18370v3 Announce Type: replace-cross Abstract: Large language models (LLMs) have demonstrated exceptional capabilities when trained within executable runtime environments, notably excelling at software engineering tasks through verified feedback loops. Yet, scalable a…
arXiv:2611.14781v1 Announce Type: new Abstract: Agents that perform state estimation using pretrained perception models, e.g., Vision-Language Models (VLMs), face the fundamental challenge that predictions are uncertain. Still, most approaches determinize these predictions into …
Video generation models produce visually realistic videos but still struggle with physical dynamics and causal consistency. Learning from generated videos requires fine-grained physical supervision, yet reliable quality judgments, correction pairs, and controlled variations are d…
Decision models answer a question with a typed output, such as a choice among options with a probability for each, so one model can choose an agent's actions, evaluate actions already taken, or score every action to teach a policy with reinforcement learning. Prior work has studi…
arXiv:2611.17814v1 Announce Type: new Abstract: With the rapid advancement of Large Language Models (LLMs), an increasing number of users are turning to LLM-powered agents for conversational shopping in e-commerce. However, the user needs continue to change as new product catego…
arXiv:2611.13006v1 Announce Type: new Abstract: Medical vision-language models (VLMs) often report a finding because the question names it; standard benchmarks miss this. From public segmentation datasets we build counterfactual views without manual annotation: the lesion erased…
arXiv:2611.10156v1 Announce Type: new Abstract: Empowered by Large Language Models (LLMs), deep research systems integrate online information into research reports through multi-step web searching and information synthesis. Evaluating these reports presents a significant challen…
Google’s Project Zero will disclose the vulnerabilities it finds 90 days after reporting them to a vendor, whether or not the vendor has shipped a fix. The deadline drops to 60 days on February 2, 2027, and to 30 days on January 3, 2028. Project Zero calls the new policy “Full X,…
🧩 Chức năng: Công cụ quanh mô hình ngôn ngữ (prompt/agent)
Open-source AI tutor that teaches on a whiteboard. Benchmark of 6 LLMs (Claude, Gemini, DeepSeek): 288 lessons drawn stroke by stroke, blind-reviewed. 白板 AI 导师 / AI 老师讲课评测
arXiv:2611.17339v1 Announce Type: new Abstract: Hallucination detection for large language models is dominated by methods evaluated on 7B-70B research models with abundant sampling budgets and exact-match or ROUGE correctness labels. The small models practitioners actually ship …
arXiv:2611.17501v1 Announce Type: new Abstract: Recommendation and Ranking (R&R) models, particularly in digital advertising, require massive inference throughput but exhibit extreme sensitivity to precision loss. Unlike Large Language Models (LLMs) that frequently tolerate quan…
arXiv:2611.17846v1 Announce Type: new Abstract: Abliteration, a fine-tuning-free weight edit, removes refusal from open-weight large language models (LLMs). The developer of Arbitrary-Rank Ablation (ARA), an optimisation-based variant, reports that direction-based abliteration, …
arXiv:2611.17377v1 Announce Type: new Abstract: Visual adversarial robustness remains a fundamental security property as vision systems evolve from conventional classifiers to vision-language models (VLMs) and large vision-language models (LVLMs), with visual inputs remaining an…