Claude discovers a novel enzyme system with CRISPR-like repeats
(anthropic.com)
316 points by raahelb 3h ago |
The Internet Right Now — Live Technology Dashboard
One live view of what tech is reading, building, and publishing across Hacker News, GitHub, and arXiv.
Hacker News
Top stories, links, and active comment threads from the Hacker News front page.
Checked just now
Once Claude can measure something, it can make it faster
(claude.dev)
106 points by matthieu_bl 2h ago |
Fixing the Portobello Police Station Clock
(pointinthecloud.com)
335 points by avidly 6h ago |
A brief history of Windows scroll bar shortcuts
(devblogs.microsoft.com)
71 points by tybulewicz 3h ago |
Italian parliament votes for return to nuclear energy
(apnews.com)
400 points by geox 4h ago |
LensVLM: Compressing long context as images, expanding only relevant pages
(huggingface.co)
7 points by victormustar 3h ago |
Gemini 3.8 text-to-speech
(blog.google)
208 points by swolpers 6h ago |
Radicle: Disclosure of Vulnerability in the Network Protocol
(radicle.dev)
97 points by lostmsu 6h ago |
Jev in 25 Lines of Python
(nobodywho.ai)
583 points by bashbjorn 14h ago |
Tokens too cheap to meter
(jyn.dev)
195 points by teoruiz 12h ago |
Popular new repos
Repositories created in the last seven days, ranked by total stars.
Checked just now
NandhaKishorM/laya
(github.com)
Non-autoregressive System 1 decision engine. Typed choice, score and yes/no decisions over any text in a single forward pass, in 100+ languages, with a router that picks the right checkpoint per request.
⭐ 20236 | Python | 5d ago | 74 issues
tamaratran/fast-jev-compaction
(github.com)
Claude Code plugin that replaces the compaction summary with Jev decisions: every tool call and result is scored in one fast request, stale ones are dropped or truncated, everything kept stays verbatim.
⭐ 6554 | TypeScript | 6d ago | 78 issues
zai-org/ZCode
(github.com)
Z.ai's coding agent harness. Powerful, intelligent, extensible.
⭐ 6525 | TypeScript | 3d ago | 11 issues
mizorewww/laya-mlx
(github.com)
Native MLX runtime for Laya typed decision models — 7–14 ms short decisions on M3 Max. No text generation, PyTorch, or cloud API.
⭐ 5920 | Python | 4d ago | 11 issues
jaredpalmer/kev
(github.com)
tiny Jev-like family of decision models built on top of Qwen3.5 you can train and run on your own
⭐ 5888 | Python | 6d ago | 24 issues
jev-chat/jev-chat-jarvis
(github.com)
装在手机上的对话副驾：在微信 / QQ / X / 飞书里读懂对方、给出候选回复、一键填入输入框，发不发由你。非侵入，只读屏幕，不 hook 不改包。
⭐ 5198 | Kotlin | 2d ago | 40 issues
TianyuCodings/NanoJev
(github.com)
A nano replica of Jev: parallel decisions, dynamic candidates, and an end-to-end training pipeline.
⭐ 2109 | Python | 6d ago | 7 issues
unreallabsai/unreal-agent
(github.com)
Async-first agent harness
⭐ 1688 | Go | 1d ago | 8 issues
bespokelabsai/nimble
(github.com)
Local typed decisions, contrastive data curation, and model evaluation.
⭐ 1679 | Python | 5d ago | 4 issues
hydra-db/open-glean
(github.com)
An open-source AI platform for knowledge work. Connect your apps, find answers, and get work done.
⭐ 1504 | TypeScript | 6d ago
arXiv Papers
New papers from cs.AI, cs.LG, cs.CL, and stat.ML.
Checked just now
Flash-dLLM: IO-Aware KV Caching and Parallel Decoding for Fast, Memory-Efficient Diffusion LLMs
(arxiv.org)
Quan Nguyen-Tri et al. | 1d ago |
A Decentralized Partially Observable Team Decision Methodology with Delayed Information Sharing
(arxiv.org)
Xiaoxing Ren et al. | 1d ago |
Agensh: Scaling Organizational Intelligence to 1,024 Agents
(arxiv.org)
Zhihao Zhan et al. | 1d ago |
SpeakerMem-R1: Speaker-Centered Dual-Track Memory for Multi-Party Dialogue
(arxiv.org)
Haobo Zheng et al. | 1d ago |
CliffCompaction: Cost-Efficient Compaction for Long-Horizon Coding Agents
(arxiv.org)
Trang Nguyen et al. | 1d ago |
SWE-Serve: Benchmarking Agentic Engineering For Production Inference Serving
(arxiv.org)
Jennifer Williams et al. | 1d ago |
A2M: Trace-Optimized Agent Hijacking in the MCP Ecosystem
(arxiv.org)
Laizhen Li et al. | 1d ago |
Grow the Harness, Not the Context: From Strategy-Free Scaffolds to Reusable Specialist Agents
(arxiv.org)
Laizhen Li et al. | 1d ago |
Type-Safe Is Not Error-Free: A Constrained Decision Head Follows the Option Name, Not the Rubric Bound to It
(arxiv.org)
Yu Sun et al. | 1d ago |
FleXray: Universal Clinical X-ray Segmentation
(arxiv.org)
Victor Ion Butoi et al. | 1d ago |