Archive
1523 technical news articles
AI 安全
GitHub has converted its lowest-cost Lite review into a multi-agent ensemble, with 47% more high-severity findings adopted in official experiments. Review agents can now also execute shell tools, making runner isolation, prompt injection, and audit logs for automatic resolution more important.
最新模型
XPeng uses behavioral probes, cross-scale distillation, and LoRA to shrink the Qwen3-ASR-0.6B audio encoder from 18 layers to 14. The weights and inference code are public, but the full distillation pipeline and more than 280,000 hours of training data have not been released.
AI 基礎設施
Together AI ported its open-source CUDA kernel framework to NVIDIA Vera Rubin, reorganizing data reuse and the memory pipeline. Tests with square 16K matrices reached 22.24 PFLOPS, though the figure comes from a qualification sample and the still-maturing CUDA 13.4.
推論系統
The new release serves MiniCPM-o 4.5, PersonaPlex, and Nemotron VoiceChat through a shared WebSocket runtime, while adding diffusion batching, parallelism, and component offloading. It remains a pre-release, and some endpoints silently fall back to turn-based voice, so a successful connection alone does not confirm full-duplex capability.
AI 開發平台
The public beta provides managed sessions, context compaction, tool orchestration, and sub-agent delegation, eliminating the need for developers to maintain agent loops themselves. Execution can take place in an OpenAI sandbox, on your own infrastructure, or through a partner provider, but versions, permissions, and usage costs must still be controlled separately.
代理工具
Cursor is expanding the unit of work for coding agents from a single conversation to a Project that can span multiple PRs and continue for months. A coordinator agent maintains shared files, delegates work to cloud and local agents, and automatically starts tasks in response to Slack messages, schedules, or PR events.
AI research
The new method arranges each recurrent-network update into a continuous flow from noise toward an answer, allowing the model to learn state across steps without fully backpropagating through every iteration. At inference time, more integration steps can be used for additional compute, while different initial noise samples can produce multiple candidate solutions.
AI 安全
GitHub has added organization- and repository-level REST APIs that let security teams programmatically deploy AI vulnerability scanning for pull requests at scale. The scans can cover languages and frameworks not supported by CodeQL, but findings are currently advisory only, consume AI credits, and cannot be added to rulesets as merge requirements.
最新模型
Sakana AI has launched Fugu Max and Fugu Ultra v2, optimizing multi-model orchestration along two separate tracks: lower cost and greater capability. Both retain an OpenAI-compatible interface, but the official benchmarks still combine Sakana’s own tests with vendor-reported results, providing insufficient evidence of cost advantages in real-world workloads.
AI 安全
Anthropic’s new threat report shows that several types of attackers have upgraded Claude from a coding assistant into a schedulable attack-orchestration layer with persistent memory. Some workflows automatically modify, rebuild, and redeploy malware after detection, although humans still largely control target selection and the review of stolen data.
最新模型
OpenAI is giving developers access to GPT‑Live‑1, which can listen, speak, and handle interruptions simultaneously, while allowing the voice front end to delegate complex reasoning and tool operations to other models. The new architecture reduces fragile handoffs between STT, LLM, and TTS, but backend reasoning, tools, and call infrastructure are billed separately.
AI for Science
ARCHE integrates literature retrieval, competing hypotheses, electronic structure calculations, and reflective review into a resumable closed loop, while releasing 65 actions and data from three case studies. However, the archived data already differ from figures reported in the paper, and the specialized model weights, training data, and some chemical evidence were not released with the code.