Adrien Brault-Lesage (@AdrienBrault) header
Adrien Brault-Lesage (@AdrienBrault) profile picture

Software Engineering, Product, LLMs, Open Source, Home Automation.

Bailly Romainvilliers, France
Joined September 2009
Photo by Adrien Brault-Lesage (@AdrienBrault)Photo by Adrien Brault-Lesage (@AdrienBrault)Photo by Adrien Brault-Lesage (@AdrienBrault)Photo by Adrien Brault-Lesage (@AdrienBrault)Photo by Adrien Brault-Lesage (@AdrienBrault)Photo by Adrien Brault-Lesage (@AdrienBrault)Photo by Adrien Brault-Lesage (@AdrienBrault)Photo by Adrien Brault-Lesage (@AdrienBrault)Photo by Adrien Brault-Lesage (@AdrienBrault)Photo by Adrien Brault-Lesage (@AdrienBrault)
Adrien Brault-Lesage retweeted
Oh my god: Anthropic mathematician Levent Alpöge (@__alpoge__) says he has proven the Riemann Hypothesis, open for 167 years. "I believe I have, working with Claude, discovered a proof of the Riemann Hypothesis." He says the method also covers the Generalized Riemann Hypothesis. The proof is out as Lean code (a proof language where a computer checks every step), so experts should be able to check it quickly, Scientific American reports. Same Alpöge who was behind this year's Jacobian conjecture result and the Navier-Stokes work with Tristan Buckmaster. I assume we will get an official statement very soon from him!!
145
315
136
3,280
193,096
Adrien Brault-Lesage retweeted
I present QuixiInfer - my ridiculously optimized inference engine for running DeepSeek-v4.1-Flash-1017 on a single GPU, 256k context. Getting ~25-30 tok/s, and up to 3 concurrent requests on 5090. Implemented from scratch in Rust, inspired by DeepInfer and Strata. Uses dynamic CPU and NVMe offloading. I would love to get contribution from people willing to test on other GPUs like 3090, 4090, AMD. @casper_hansen_ @AlpinDale
46
29
4
385
14,064
Adrien Brault-Lesage retweeted
Finally some good news !! About Qwen4 ofcourse ! Qwen dev's have been improving Qwen4 so much, thanks to RL that the version we are going to get is a vastly improved version, basically a version bumb or aka Qwen4.1. This is really good, they know they can't release an inferior model, pressure is huge. Now my hopes for Qwen4-27B is even higher !! LFG !!!! Only question is when, but they can take a bit more time if it means we get something that meets our expectations !
btw, qwen4 is a whole different model now from when i first used it like 2 months ago now. it's insane its been this long of a release timeline. it's basically qwen4.1 now with all the extra RL they've been doing
25
19
4
504
31,077
Adrien Brault-Lesage retweeted
Breaking: 35 premium tools free for a year—now with OpenRouter, LangChain, Sentry, TextExpander, and Ploy 💥 I'm excited to welcome five incredible new partners to the Lenny's Newsletter Product Pass: 1. @OpenRouter ($650 value, Insider only): One API and one bill for hundreds of AI models. 2. @PloyAI ($600 value): AI agents that run growth marketing for your website. 3. @Sentry ($360 value): Know the moment something you shipped breaks, and fix it fast. 4. @TextExpander ($300 value): Your best prompts, replies, and phrases, always just a few keystrokes away. 5. LangSmith by @LangChain ($468 value): See what your AI agents are actually doing, and make them better. This brings the LennysProductPass.com bundle to 35 products, worth $41,428 (!!) for Insider subscribers ($400/year), and 22 products worth $37,062 for Annual subscribers (just $200/year). This bundle includes a full free year of @WisprFlow, @NotionHQ, @Lovable, @Linear, @Higgsfield, @Replit, @ElevenLabs, @RunwayML, @Railway, @Resend, @Granola, the @X Subscription (with Grok Bot, Cursor, and X), and so much more—all included in your Lenny's Newsletter subscription. I'm pretty sure this is the greatest deal in the history of software. If you are already an Annual or Insider subscriber, grab your codes here: lennysproductpass.com/ If you're currently on a free or monthly plan, upgrade to immediately unlock the Product Pass: lennysnewsletter.com/subscri… Codes are limited and first-come, first-served. The most popular products can sell out, so grab yours now. Learn more here: lennysnewsletter.com/p/35-pr…
38
12
22
312
74,300
Adrien Brault-Lesage retweeted
Today's the day. OVH is releasing Olala mini 7A1B, the model that kicks off its very first family of language models. A compact 7-billion-parameter model, trained on European public supercomputers. It is built on a brand-new architecture, runs even without a GPU, and is released as open source, available on Hugging Face. Learn more 👉 llm.ovhai.com/olala-mini-v1-… #OVHcloudAISummit @OVHaiLLM
14
42
10
161
20,194
Adrien Brault-Lesage retweeted
Never before has this amount of power sat quietly under a desk.
34
17
6
533
20,362
Adrien Brault-Lesage retweeted
GPT-6.2-Astra on SimpleBench what is OpenAI doing
65
15
3
644
41,483
Adrien Brault-Lesage retweeted
We just made LLM inference up to 30% faster ⚡ Introducing DNeo, a new diffusion drafter designed in collaboration with @cerebras researchers. The secret? Your draft model needs a working memory, and the "right" context encoding mechanism. 📄 Paper: alphaxiv.org/pdf/2611.21925
8
36
20
176
19,062
Adrien Brault-Lesage retweeted
Gemini 4 Neon: frontier-level intelligence with fast end-to-end task completion across complex software engineering, knowledge work, and science workflows. Begins rolling out today in the Gemini API in @GoogleAIStudio, Gemini for business, for paid Google AI Pro and Ultra subscribers in Google @Antigravity, and for Google AI Pro and Ultra subscribers in the @GeminiApp.
912
1,479
681
18,364
2,094,189
Adrien Brault-Lesage retweeted
vLLM v0.34.0 is live! 🚀 1,061 commits · 412 contributors 🤖 10 new models 🗄️ Extensible KV cache on by default 🏃 Model Runner V2 is now the only GPU runner ⚡ Up to 1.95× faster Step5 NVFP4 decode on 4× GB200 ⚠️ Upgrade notes github.com/vllm-project/vllm…
18
28
9
206
15,037
Adrien Brault-Lesage retweeted
It's been nearly 6 months since I posted a 'watch me work' deep dive on my skills. Since then, they've become the most-starred skills repo in the world. So, here's 70 minutes of me building a 21,000-line PR, especially focusing on: - DDD with /grill-with-docs - Improving your codebase with /retro - How to do human review effectively Enjoy!
113
304
36
4,969
324,623
Adrien Brault-Lesage retweeted
How much can you improve a small model in 18 days on 144 GPUs? Meet DaisyRose-27B, RadixArk's first open-weight model! It is trained with RL on Qwen3.8-27B, with 1,200 curated long-horizon agent tasks for 18 days on 144 H200s. Powered by Miles and our own training backend, we stretched agent sessions up to 1M tokens and raised the Terminal-Bench 4.1 score from 5.6% to 12.0%. See how we pull it off ⬇️
59
116
62
812
556,292
Adrien Brault-Lesage retweeted
Introducing Firecrawl Local. We rebuilt the Firecrawl repo so agents can scrape locally. - 37x faster rust based scrape primitive (vs self-hosted) - 230x less memory (11 MB idle) - 100% open source Compatible out of the box with all of our MCP & SDKs. github.com/firecrawl/firecra…
235
340
121
5,181
765,278
Adrien Brault-Lesage retweeted
"Instead of trying to control the whole world, we go directly to the AI companies." They're saying the quiet part out loud now...
Jack Clarke of Anthropic was reportedly the one who leaked the Jensen Huang spy story to The New York Times. The New York Times is controlled by the Jewish Ochs-Sulzberger family through the Ochs-Sulzberger Trust. As of March 2026, the trust holds approximately 94.6% of the Class B stock. During Jensen Huang's recent October tour he publicly voiced support for open weight, locally run AI models, which was a direct threat to Anthropic and OpenAI. This was also a threat to Israel's control over the entire AI industry, since they have their people in place at the head of Antropic, OpenAI and Palantir. The media is already preparing the public by announcing "Meet the 3 Israelis who could inherit Nvidia's AI empire" This was an Israeli psyop to oust Jensen from $NVDA and take control of the one company that threatened their control over the entire AI industry.
94
914
62
3,603
117,160
Adrien Brault-Lesage retweeted
F*CK IT $10 for $500 GLM 5.3 till end of the week
we're bored we're just gonna do phase 2 now $200 of GLM 5.3 usage for $10 have fun
747
519
551
10,539
917,271
‼️ UPDATE: Security firm Calif says it took over Jira, Confluence, Jenkins and Keycloak servers with Java4Shell, without even logging in. The flaw sits in the code Java uses to decode Base64, a common way to send data as text. A short malformed string makes it write past the end of its memory. Calif's researcher says OpenAI's Codex helped him find the bug in an hour and build a working exploit in two more, for under $200 in AI costs. Oracle shipped a patch tonight. Calif is holding back its exploits until people have patched.
🚨 BREAKING: A critical Java flaw that reportedly lets hackers hijack servers without a password leaked before its patch. Security firm Calif says it found the bug, dubbed Java4Shell, and will publish details today. Oracle fixes 11 Java flaws today but hasn't tied any to it.
56
449
88
3,265
303,080
Adrien Brault-Lesage retweeted
Java4Shell: write once, RCE everywhere calif.io/java4shell
59
694
144
2,671
231,972
Adrien Brault-Lesage retweeted
Check out Code Mode, a new, experimental way for Claude Code to work with tools. Claude writes scripts that chain your tools together in sophisticated ways, and it can answer you while long commands are still running. Opt in with CLAUDE_CODE_EXPERIMENTAL_CODE_MODE=1 code.claude.com/docs/en/code…
451
771
369
12,349
2,304,327
Adrien Brault-Lesage retweeted
while everyones focused on funsies skills and shit the most useful thing you can do to accelerate your company is to make LLM generated artifacts easier to understand whether thats a review walkthrough (read: not a mermaid diagram or more markdown text), or better visuals to explain things, thats whats needed to go faster and have trust
39
11
4
290
15,408