“编织者”啊， 这样看，Anthropic 的股票代码是 ANT 更有意思了 😆
We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math.
It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training.
We call this Weight-Space Programming: frontier LLMs can now natively read/write neural network weights! 🧵
Blog: rikiyat.github.io/blog/weigh…
Readers added context they thought people might want to know
LLM was given weights explicitly as text in prompt and used code/forward passes (per authors' logs) to ablate Qwen looping direction, a known thinking-mode issue per model card. Primarily fixed repetition in that mode, not general 'natively' improving weights/math.
rikiyat.github.io/blog/weight-sp…
github.com/RikiyaT/lm_und…
huggingface.co/Qwen/Qwen3.5-0…
huggingface.co/RikiyaT/Qwen3.…
unclebill retweeted
SEO / GEO 小帖士：URL 里用 -，别用 _。
Google 10/23 在 Search Central 文档里正式写明：连字符会被当成分词符，下划线不会。
也就是说 seo-tips 是两个词，seo_tips 在 Google 眼里是一个词。
你们的站还有下划线 URL 吗？
下划线基本上都是老程序员带出来的东西。
传说中url里带词，会对排名有帮助。这个是肯定的，我很确定Bing是有帮助的，Google 肯定也是有帮助的。
下划线这个东西么。。做过开发的都懂的。。
但是请
把这个文档甩给开发。
没有什么费力啊，改起来要花时间啊，性价比不高这种事情。你问我值不值得做？要做就把基础的，细节全部都做好了。
SEO就是积少成多的，就是细节，都是开卷考。
你细节没做好，这里扣一分，那里扣一分的。
老板，甲方问下来。。锅谁背？
这次的连接是英文版的，中文版的翻译没更新 lol
developers.google.com/search…
unclebill retweeted
tsrs is super fast, also for Hono. I benchmarked with the script we always use for Hono’s TypeScript types
11x faster than tsc, 4x faster than TypeScript 7.0.x, 3x faster that 7.1.x
unclebill retweeted
Imagine your modern software pauses to garbage collect while you're in the middle of Doing a Barrel Roll in your car.
unclebill retweeted
Oh my God! Was working just found out about RJ
I am heartbroken
Though RJ didn’t play my TV son, I always looked out for him as a young kid on set. He was a sweetheart of a young man and will be greatly missed. This is so heartbreaking, I can’t believe it. Thinking of him and his family.
💔💔💔💔💔💔
unclebill retweeted
小西天，看着像视频，但其实是我们在现场实地拍摄 3,811 张 206 GB 的照片后建模的。FUNES 把《黑神话：悟空》里「既见未来，为何不拜」满天神佛的原型，来自自山西临汾隰县的小西天，做成了一个可漫游的 3DGS 数字存档。
完全实地拍摄，每天清晨一开门就冲上山去，趁着没人的时候拍。然后通过 Gaussian Splatting 重建，没有手工建模，尽量保留真实悬塑和圆塑的极其密集的金色空间、细节和光感。不同的材质在这里交织成了无法分辨的一个天国世界。这种半空中的小塑像是「悬塑」，它们大多出现在十六世纪到十七世纪。
在现场如果要看清小西天的所有细节，我想大概需要三天时间。但是有了模型，我们可以在屏幕前慢慢看。在相当长的时间里，学术界并没有特别重视小西天这样的悬塑——因为在只有学术图录的年代，平面印刷无法展示出悬塑的震撼。而随着技术的进步，我们终于可以在远方一窥明代悬塑的璀璨。
重轻特意为这个模型做了配乐，大家可以打开慢慢欣赏。
推荐电脑访问：funes.world/apps/the-hanging…
Hi friends, I just created a token Lucky Drop. Download Mask.io to claim.Follow @masknetwork for Web3 updates and insights.
🧧🧧🧧 Try sending Lucky Drop to your friends with Mask.io.
mask.io/?PostData_v2=AJcAq3R…
unclebill retweeted
You built a Chrome extension using an AI agent. Now, how do you let that agent test it?
We updated Chrome DevTools for agents so your agent can install extensions & interact with content inside popups automatically → goo.gle/4cjO2wa
终于回归了！
翻了下 GitHub，一大波 commit 👍
coc.nvim 的开发进展：
使用 OCR 开源工具加 GPT5.6 对代码和文档进行了完整审计，所有错误和遗漏（包括 wiki 页面）均已修复。
LSP 3.18 几乎所有特性均已被支持。
插件的集成测试工具现已发布 github.com/neoclide/coc-test ，后续将接入更多 coc.nvim 插件
MCP 服务支持 github.com/neoclide/coc.nvim…
今天更新了，效果真的不错
Next.js 16.3 is now available!
• Up to 90% less memory in dev
• Faster builds, type checking, and rendering
• Better tooling for AI agents
• Custom error boundaries
• Instant Navigations for SPA-like responsiveness
nextjs.org/blog/next-16-3
Here's what's new ↓
