👉 Live site: https://jmk9.github.io/rnews/
Daily picks in robot learning — RL / IL / VLA across cobot arms, bimanual, mobile manipulators, and humanoids. Plus an industry news feed from ~18 sources (Robohub, IEEE Spectrum, TechCrunch, NVIDIA, DeepMind, MIT, 로봇신문, AI타임스, …).
Each card carries a short LLM-written summary — read the gist in place, click the title only when you want the original paper / repo / article.
Updated every morning, 07:00–09:00 KST (new items land around 07:00; the AI summaries finish filling in by ~09:00). Sorted by what people actually use — GitHub stars, code availability, real-robot experiments.
Open https://jmk9.github.io/rnews/ in any browser. The page is a single scrollable feed, organized top to bottom:
- Code & repos — GitHub projects + papers that released code. The actionable stuff. Sorted by popularity (stars + forks).
- News & articles — fresh industry pulse from RSS feeds.
- Papers only — research papers without code. Lower priority but kept for completeness.
- Search box — full-text over title, summary, and tags (× to clear; matches highlighted in titles).
- Source:
News/arXiv/GitHub - Method:
#RL/#IL/#VLA(VLA also covers robot foundation models) - Platform:
#Manipulator/#Bimanual/#MobileManipulator/#Humanoid/#Other(no specific platform) - More filters (collapsed): Priority
High/Mid/Low· Time last1w/1m/3m· Sort byScore(relevance, default) /Created/Pushed/Stars
Filters stack — #VLA + #Humanoid shows only VLA work on humanoids.
Click All in a row to clear that dimension. The colored bar on each card's
left edge is its priority (🟢 High / 🟠 Mid / ⚪ Low).
The 2–3 sentence blurb on each card is an LLM-written summary of the source, meant for a quick gist. It can occasionally be off — open the original via the card title before relying on details.