目前没有足够证据断言 Grokbot 已经凉了，但现有观察与“正在变凉”的假设相符。
需要明确的是，“说话像 GPT”并不能推出“后端就是 GPT”。这是两个相关但不同的命题，不宜混为一谈。
现有证据能够支持的较窄结论是：我不需要两份 GPT 订阅。
我正在考虑退订 X 和 决定，而是对我是否愿意继续为这坨东西付钱的阶段性重新评估。
需要明确的是，“说话像 GPT”并不能推出“后端就是 GPT”。这是两个相关但不同的命题，不宜混为一谈。
现有证据能够支持的较窄结论是：我不需要两份 GPT 订阅。
我正在考虑退订 X 和 决定，而是对我是否愿意继续为这坨东西付钱的阶段性重新评估。
1
0
0
108
Dango233 retweeted
Energy-Based Models can now scale better than Diffusion models.
Introducing TEM: Transformer Energy-based Models
TLDR:
• TEM outscales Diffusion without guidance
• First scalable contrastive Energy-Based Models
• Better generalization, test-time scaling, and OOD detection than Diffusion
🧵
Introducing TEM: Transformer Energy-based Models
TLDR:
• TEM outscales Diffusion without guidance
• First scalable contrastive Energy-Based Models
• Better generalization, test-time scaling, and OOD detection than Diffusion
🧵
46
140
1,270
101,254
Wait, what's going on with this second image!?
🚀 DeepSeek-V4.1-Flash-1017 is here!
Stronger benchmark performance, faster task completion, and shorter outputs than the previous version—a Pareto improvement in performance and cost. Now live on our website, app, and API.
🛠️ Optimized for @DeepSeekHarness, Tencent CodeBuddy, grade to DeepSeek Harness v0.2.1-rc.1 or later for the best experience with the new model.
1/4
0
0
2
112
Dango233 retweeted
A 30M-parameter model, on CPUs, inside Postgres
It matches a 0.6B dense embedding model on top-100 recall, to within 0.004, and finds more in the top 1,000
Meet Evoke. Open source today.
It matches a 0.6B dense embedding model on top-100 recall, to within 0.004, and finds more in the top 1,000
Meet Evoke. Open source today.
20
58
518
36,834
It looks like the GPT-6 Pro in the Chat mode had some issues over the last few hours. All the conversations I started failed to return results properly, just getting stuck on Thinking and ending.
Not sure if @thsottiaux can cast his magic on chat limits...
Not sure if @thsottiaux can cast his magic on chat limits...
0
0
1
546
Dango233 retweeted
Most AI tools only do one part of building a product.
One researches.
One designs.
One writes code.
One generates images.
One deploys.
You still have to connect everything yourself.
II-Agent is different.
Give it an idea, a reference website, or an existing repo.
It ity, plans the product, builds the frontend + backend, connects databases & payments, tests it, deploys it — and keeps developing with you.
Don’t just generate code.
Build the product.
Watch it happen ↓
agent.ii.inc
One researches.
One designs.
One writes code.
One generates images.
One deploys.
You still have to connect everything yourself.
II-Agent is different.
Give it an idea, a reference website, or an existing repo.
It ity, plans the product, builds the frontend + backend, connects databases & payments, tests it, deploys it — and keeps developing with you.
Don’t just generate code.
Build the product.
Watch it happen ↓
agent.ii.inc
8
18
35
6,261
Well... What should I say? Sol is better than fable in almost all my cases lol
Or… what if we gave you $100 in Codex credits if you tell us what you love about GPT-5.6 Sol or why you switched?
Tweet it, claim your gift, enjoy more usage. First 10k get the free tokens!
switch-to-codex.openai.chatgpt.site
1
0
2
407
I really miss those days when I used to write strange cutout algorithms to express ideas
First animation from integrated PyTTI C++/CUDA port.
I used DD style (@dango233max) CLIP cutouts since that's what I already had in the codebase.. and I think that may be good.
0
0
3
306
Dango233 retweeted
Make your agent smarter.
The II-Commons skill gives your agent reliable knowledge from arxiv, PubMed & more, plug it in
Repo: github.com/Intelligent-In…
Add it to II-Agent: agent.ii.inc
The II-Commons skill gives your agent reliable knowledge from arxiv, PubMed & more, plug it in
Repo: github.com/Intelligent-In…
Add it to II-Agent: agent.ii.inc
3
12
55
8,459
Dango233 retweeted
20 years ago, my first startup was all about enterprise search. Two decades later, we’re still building search engines. The technology has shifted from NLP to NN and the users from humans to agents. but searching is still the core. opensource the fastest bm25 engine:
so we built psql_bm25s.
exact BM25 retrieval. native Postgres access method. ~23x faster than pg_search on the standard benchmark.
retrieval stops being a budget item. the harness stops rationing. the agent gets to look things up like it should have the whole time.
5
8
53
20,535
Dango233 retweeted
我們開源了這顆星球🌎上速度最快的低成本 bm25 引擎。
so we built psql_bm25s.
exact BM25 retrieval. native Postgres access method. ~23x faster than pg_search on the standard benchmark.
retrieval stops being a budget item. the harness stops rationing. the agent gets to look things up like it should have the whole time.
6
31
230
46,329
DS4 is geart!
I made a temporary fork with my weekend patches while the PRs are under review
- unlock q4 on 192GB MAC
- llama.cpp-style raw completions endpoint: enable Pre-filling and custom templates in SillyTaverns etc.
Pre-merge convenience fork only :)
I made a temporary fork with my weekend patches while the PRs are under review
- unlock q4 on 192GB MAC
- llama.cpp-style raw completions endpoint: enable Pre-filling and custom templates in SillyTaverns etc.
Pre-merge convenience fork only :)
Welcome to DS4, a specialized inference engine for DeepSeek v4 Flash. github.com/antirez/ds4
This project would have been impossible without the existence of llama.cpp and GGML and the work of @ggerganov and all the other contributors. Thanks!
1
0
3
318