Pai 的 AI 雷达
今天的简报集中呈现了 AI Agent 从学习路径、编程插件、权限治理到遗留代码迁移的工程进展，同时补充了 CodeQL 的 ARM64 与语言覆盖更新、多模态推理拆分、时序模型、系统工程上下文建模，以及 OpenAI、Anthropic、Google 的组织、政策和产品信息。阅读时可先把前段当作可直接验证的工程与平台变化，再按自己的工作场景挑选模型、治理、创作或应用方向，后段条目则适合作为按需检索的专题线索。
Agentic autofix is now available to help you burn down findings in your code quality backlog. You can select up to 25 standard findings on a page and assign the whole set to Copilot in one action. Copilot starts fixing them agentically on a branch, validates its own changes, then
arXiv:2607.12659v5 Announce Type: replace-cross Abstract: Vision-Language-Action (VLA) models have achieved impressive performance on diverse embodied tasks. However, deploying VLA models on low-power onboard devices, such as the Jetson Orin, remains challenging due to their high
arXiv:2609.05593v1 Announce Type: cross Abstract: Autonomous robot navigation failures differ not only in categorical severity but also in the physical context in which they occur. A near-miss at low speed under reliable sensing is not equivalent to the same event during rapid mo
arXiv:2609.07350v1 Announce Type: cross Abstract: Climate change is increasing the severity and unpredictability of natural disasters. In time-critical crises such as wildfires, traditional monitoring practices remain limited by coverage, cost, and personnel risk, paving the way
arXiv:2609.05927v1 Announce Type: new Abstract: Robot manipulation foundation models require scalable evaluation and data generation across diverse scenarios, with simulation providing an environment for both. Automated scene generation offers a promising path, yet prior work has
More GitHub Enterprise Cloud customers can now start a self-serve GitHub Advanced Security trial to evaluate GitHub Code Security and GitHub Secret Protection. Eligibility has expanded from enterprises with up to 100 licenses to enterprises with up to 300 licenses. To set up a Gi
Repository rulesets allow you to easily add scalable protections across your repositories. Starting today, you can use repository rulesets to block pull requests from merging when the pull request introduces secret scanning alerts. What’s new You can enable the new rule require s
Traditional software delivery assumes a static paradigm: code is constructed prior to execution and deployed as a fixed artifact. We present Agentic Just-In-Time Software Construction (A-JIT), a paradigm that replaces static binaries with dynamic, software systems that can perpet
AI agents are increasingly involved in network automation, where they can initiate configuration changes through mediated operational interfaces and assess the resulting state. Nonetheless, operational networks usually span many devices and administrative domains. Realizing an op
Today, we're releasing a demo of WeWorm, the first zero-click worm to spread through WeChat calls across iOS and Android. [...] The victim does not need to answer the call, or interact with their phone at all. Even if they do answer, they hear nothing, and the exploit still succe
Discover how filmmakers and Google DeepMind used AI to recreate a couple's unrecorded past in the short film "Love, Rendered."
arXiv:2608.07346v4 Announce Type: replace Abstract: With the rapid advancement of large language models (LLMs), harnesses have become essential infrastructure for deploying agents across a wide range of domains. The fast-evolving harness ecosystem has also made rigorous capabilit
arXiv:2609.08592v1 Announce Type: new Abstract: Evaluating tool-augmented LLM agents requires diverse, realistic user inputs yet most evaluation frameworks use flat role descriptions ("you are an angry customer") that produce near-identical conversations regardless of the underly
arXiv:2609.06914v1 Announce Type: new Abstract: Large language models (LLMs) have demonstrated strong capabilities across diverse domains, showing considerable potential in medicine. However, their application in medical settings remains limited by the scarcity of visual question
arXiv:2609.05837v1 Announce Type: new Abstract: LLM-based agents are increasingly deployed in real-world applications through tool-use APIs, yet training them for specific environments remains fundamentally difficult: real-world applications provide no pre-defined tasks or verifi
arXiv:2609.05818v1 Announce Type: new Abstract: We introduce ABLE, a benchmark for evaluating LLM agents' ability to use biological AI models (BAIMs), such as ProteinMPNN and AlphaFold3, in dual-use protein design workflows. ABLE assesses agent performance through a set of tasks
arXiv:2606.19795v2 Announce Type: replace-cross Abstract: Electronic design automation (EDA) is multi-stage and handoff-heavy, relying on transfers among humans, design artifacts, and multiple tools. LLM-based agents now participate in these transfers, yet the resulting research
arXiv:2608.16690v2 Announce Type: replace Abstract: Multimodal large language models (MLLMs) are widely used for automated annotation, yet their per-class accuracy varies widely (e.g., 12%-98% across the 13 classes of three classroom sub-datasets) and is expensive to measure: eva
arXiv:2609.06879v1 Announce Type: cross Abstract: Steering vectors have rapidly emerged as a popular and effective method for guiding the output of LLMs in very specific ways. But constructing accurate steering vectors is a difficult manual process due to the opacity of embedding
arXiv:2609.05794v1 Announce Type: cross Abstract: Open-weight large language models face a low-cost white-box threat from representation engineering attacks. Attackers can estimate refusal directions and search for projection-matrix edits that suppress safety alignment while pres
arXiv:2609.05736v1 Announce Type: new Abstract: LLM tool agents can be improved without retraining by modifying the runtime harness around a fixed model: prompts, tool interfaces, middleware, state handling, and recovery logic. We study this setting as resource-bounded harness se
arXiv:2608.16207v2 Announce Type: replace Abstract: Consider a firm that surveys its competition for a particular agentic task and seeks to offer superior accuracy at every price point. A firm that Pareto-dominated its competitors would leave no rational customer a reason to buy
arXiv:2606.19376v2 Announce Type: replace-cross Abstract: Inference costs for large language model (LLM) applications are rapidly growing, driven by surging demand and rising infrastructure cost. Users expect high-quality responses, and in commercial settings this is formally cod
arXiv:2609.07247v1 Announce Type: new Abstract: Scaling the interaction horizon-the maximum number of environment interactions per episode-improves LLM agents on long-horizon tasks, and curriculum-based methods that progressively expand the horizon outperform fixed-horizon altern