Python services for the P1 smart meter pipeline. The decoder subscribes to raw DSMR telegrams from MQTT, parses them, and publishes structured JSON readings for OpenHAB. This package lives inside the ...

Exploratory code for working with open human ECoG speech data. The current focus is DANDI 000019, a 256-channel ECoG dataset collected during consonant-vowel syllable production. The first version of ...

AI found 21 FFmpeg zero-days, some 20 years old; Chrome 149 patched 429 bugs, including 100+ critical/high flaws.

完全跑偏的那一半：前端工作的天花板，不是切页面。 前端真正值钱的能力——异步流程设计、流式体验优化、交互状态管理、组件化工程思维——这些 AI 一个都学不会。而这些能力，恰好是 AI Agent 应用开发最核心的竞争力。

过去一年，开源模型的发布节奏已经快到让人麻木。每次发布，伴随的永远是一组跑分、一张能力雷达图，以及几个“超越某某”的结论。

2026 年 6 月 1 日，国际机器人与自动化会议（ICRA）在奥地利维也纳召开。次日上午的自动驾驶与导航报告环节，雷峰网GAIR 2021大会嘉宾、上海交通大学教授王贺升发表了题为《Learning to Navigate: From Scene ...

Google DeepMind just rolled out Gemma 4 12B, a 12-billion-parameter model that can parse text, images, audio, and video ...

Google's Gemma 4 12B brings multimodal AI — audio, video, and text — to a standard 16GB laptop in 2026. No cloud required. Here's what it does and why it matters.