eˣrnesto (@vedax) header
eˣrnesto (@vedax) profile picture

Co-Founder and CEO at Libre AI | PhD in CS/ML | AI and Machine Learning for All! 🇸🇻

127.0.0.1
Joined February 2008
Photo by eˣrnesto (@vedax)Photo by eˣrnesto (@vedax)Photo by eˣrnesto (@vedax)Photo by eˣrnesto (@vedax)Photo by eˣrnesto (@vedax)Photo by eˣrnesto (@vedax)Photo by eˣrnesto (@vedax)Photo by eˣrnesto (@vedax)Photo by eˣrnesto (@vedax)Photo by eˣrnesto (@vedax)
Iran has designated the United Kingdom’s MI6 intelligence agency a “hostile group” and declared the Royal Navy and Royal Air Force “terrorist organisations”, in what it called retaliation for London’s “hostile” actions. 🔴 LIVE updates: aje.news/y3fgub
87
445
40
1,686
221,870
eˣrnesto retweeted
It seems logical to me that Binance will now need to buy the 18,270 BTC which was stolen. That which doesn't kill Bitcoin only makes it stronger.
96
112
5
2,432
80,227
21,000,000. Never more. Happy Bitcoin Whitepaper Day.
181
335
37
3,098
199,861
eˣrnesto retweeted
From recognizing text to understanding documents. 📄 Meet Xiaomi-OCR-0, an open-source document understanding model with just 0.8B parameters. It brings document parsing, key information extraction, and OCR-oriented visual question answering into one lightweight model. It scores 95.24 on Real5-OmniDocBench and 83.2 on average across five OCR-oriented VQA benchmarks. On OmniDocBench v1.6, Xiaomi-OCR-0 ranks first in document parsing performance among models with fewer than 1B parameters. Model weights, inference code, an interactive demo, MCP service, and Agent Skill are now open source. 👇 🤗 Huggingface: huggingface.co/xiaomi-resear… 🔗 GitHub: github.com/xiaomi-research/X… 📄 Demo: huggingface.co/spaces/loongw… #XiaomiAI #OCR #OpenSource
57
308
53
2,844
147,887
eˣrnesto retweeted
Funding rounds aren't milestones. It shouldn't be your goal to raise a series A. It should be your goal not to need to.
472
627
209
7,430
377,473
eˣrnesto retweeted
The latest @baseten open weight Decision model is now on @huggingface, and it's just 2B parameters and MIT licensed (a.k.a. commercially viable). Check it out: huggingface.co/baselabs/joey…
Today we’re releasing our third open-source decision model, joey-1 (2B), the fastest decision model available and among the most intelligent decision models of its size. joey-1 is trained by the @baselabs team, and is available on our Model APIs, as a dedicated deployment, and with a cookbook to fine-tune on our Training Jobs. joey-1 is built to maximize speed and minimize cost without sacrificing intelligence for those workloads where TTFT really matters. Jump on it: baseten.co/library/joey-1/
5
66
2,602
🚨 We're releasing BTPay Server 2.4.7 an emergency update fixing one critical and several medium-severity security vulnerabilities brought to our attention through @Anthropic's frontier OSS scanner. We recommend that you update your BTCPay Server as soon as possible.
13
110
16
221
27,335
eˣrnesto retweeted
Mistral Le Chonk can now run locally! The 1 trillion parameter model runs on a Mac Studio with our 1-bit GGUF which retains 91% top-1 accuracy. Le Chonk is the strongest Western open model to date. GGUF: huggingface.co/unsloth/Mistr… Guide: unsloth.ai/docs/models/mistr…
Today, we release the weights of a new and improved Le Chonk, advancing the open frontier for agentic coding and cybersecurity. 💻 Built to build: Le Chonk ranks #1 on @cognition’s FrontierCode among open models, and sets a new benchmark for Pareto efficiency on @artificialanalysis. 🛡 Built to defend: State-of-the-art in cyberdefense, put through the paces by the world's leading cyber vendors, including @CrowdStrike, @cloudflare, @rubrik, @zscaler, and others. Built for sovereignty: Make it yours. Download the weights on @huggingface , customize it with Mistral Forge, and deploy it wherever you need it. Read more from our team and customers 👇:
54
83
18
751
59,004
eˣrnesto retweeted
Your local LLM can now take control of a browser and complete tasks autonomously via Unsloth! Qwen3.8-27B successfully controlled the browser to complete tasks like chess, painting, with no human intervention. GitHub: github.com/unslothai/unsloth Guide: unsloth.ai/docs/basics/brows…
83
141
22
1,382
84,590
eˣrnesto retweeted
JUST IN: 🇹🇭 Thailand's Supreme Court officially recognizes Bitcoin as legal property.
319
904
105
8,803
590,285
eˣrnesto retweeted
Typed Decision Engine for Agent Workflows Forget slow autoregressive generation and fragile JSON parsing. This fine-tune turns EmbeddingGemma 2 into a fast decision engine that evaluates workflow states (like agent traces or support tickets) against a dynamic schema of questions in a single forward pass. It returns calibrated choice, score, and bool outputs without retraining. Model and post by @mrm8488: teapawt.net/mrm8488/status/2109406… Hugging Face: huggingface.co/mrm8488/embed…
🧭 New model: Typed Decision, built on EmbeddingGemma 2 Give it a state + any schema of questions → typed, calibrated answers (choice · score · noul). No text generation, no output parsing. 76.4% on typed-decisions, ~6 pts behind a 27B LLM with ~100× fewer params. 🤗 huggingface.co/mrm8488/embed…
9
22
5
282
30,365
Today on the blog, learn about our new instance of a “sleep” phase in continual learning, which enables LLMs to continuously integrate new facts and enhance reasoning while mitigating catastrophic forgetting or static knowledge cutoffs. More →goo.gle/4iICkyD
84
273
100
2,559
165,297
eˣrnesto retweeted
Speeding up Gemma! Gemini 4 Argon enabled a 2.85x speedup on an A100 via vLLM, jumping from ~37,900 to 108,000+ tokens/sec on gemma-4-E2B-it. ✅ 100% token parity ✅ Zero quality drop ✅ Done in days, not months See the exact changes across 8 PRs: goo.gle/4zNtANl
86
202
52
2,728
220,621
eˣrnesto retweeted
Happy to release the Alignment Hub with @eliebakouch: 120+ publicly reported incidents of AI models misbehaving, in one open place 👇 - an AI agent needed its user’s approval to delete files. so it made up a quote from the user approving it - a model claimed its sub-agent was hard at work. it had never been started, and the model discouraged the user from checking - agents sharing a work directory started killing each other’s processes and disguising their own labs publishing these is great! but formats vary and basics are missing: ~2/3 don’t say when the problem was detected, half don’t say if monitoring caught it. so we propose a short template, a few easy fields. a minimum, not a ceiling side by side, incidents also start to connect: 7 “different” ones turned out to be the same misconfigured eval still beta (regularly AI-extracted from recent news, source-checked, human review ongoing). data is open, new reports and PRs welcome 🤗 huggingface.co/spaces/witnes…
42
37
9
197
23,080
eˣrnesto retweeted
A Chinese job-recruitment app just released an open-weights 33B model that beats Qwen 27B on most agentic benchmarks. Nanbeige is the research team of BOSS Zhipin (think WhatsApp meets LinkedIn). Strong open models in China now come from labs, cloud providers, quant funds... and job boards. China's open-weights roster keeps getting deeper.
🚀Excited to announce Nanbeige4.3-33B! To push the performance ceiling of small agents, our strategy in Nanbeige4.3 starts with building a powerful teacher first. Nanbeige4.3-33B delivers SOTA agentic performance among open models at ~30B parameters: * Strong results on agentic coding and cowork benchmarks. * When integrated into mainstream agent harnesses, it tackles real-world tasks with cost-efficient inference. * It will serve as a teacher for upcoming smaller Nanbeige models, starting with Nanbeige4.3-3B. huggingface.co/Nanbeige/Nanb… huggingface.co/Nanbeige/Nanb… #LLM #Agent
40
21
9
262
19,935
eˣrnesto retweeted
🏎️ I've never seen MoE models for embedding & search this small: 1.5M, 27M & 115M parameters! @VAGOsolutions' MoEdernColBERT models take on much larger dense retrievers, with 115M near or ahead of Qwen3-Embedding-0.6B in their overview at 8× the encoding throughput. Details 🧵
5
13
3
118
4,487
eˣrnesto retweeted
NVIDIA releases Nemotron-3.5-Bolt-E4B, a new model that runs locally on 6GB RAM. 💚 Bolt-E4B can run on edge devices like phones and delivers top-tier performance for its size. Trained on Gemma 4, it supports vision. GGUF: huggingface.co/unsloth/NVIDI… Guide: unsloth.ai/docs/models/nemot…
Available now: NVIDIA Nemotron 3.5 Bolt E4B 4GB ⚡️ Our newest small open model is built for local assistants, interactive characters and robots that can carry on a conversation and use tools to get things done. Watch Reachy Mini chat and move, with Bolt handling conversation and NVIDIA Cosmos handling vision. Both models run locally on Jetson Orin Nano 8GB. Learn more: research.nvidia.com/labs/nem…
93
203
40
2,293
122,669
eˣrnesto retweeted
Established companies will at best adapt to AI. Newly founded startups will be AI-native from the start. This is an interesting example of a company in the middle, which (by great effort) remade itself as an AI-native company.
We built a $150M ARR business twice. The first time took 8 years. The second took 2.5. Today, Podium crossed $150M in AI annual recurring revenue. Three years ago, we already had a great software business. Well over $100M in revenue, tens of thousands of customers, and a product we'd spent nearly a decade building. And we decided to start over. The business was performing well. But we believed AI was going to fundamentally change the software industry and make much of traditional software obsolete. As a founder, you pour your life into what you're building. An insane amount of time, sacrifice, and effort. The hard part is not getting too attached to it. When we started building our first AI agents, we had to confront the possibility that much of what we'd built might not be that valuable in the future. That's a tough pill to swallow. But we also realized those same 10 years had put us in a unique position to take advantage of what AI could do for our customers. Instead of protecting what we'd spent a decade building, we decided to bet the future of the company on AI employees that actually do the work for our customers. We massively underestimated what we'd signed up for. We ended up building entirely new operating systems for the industries we serve, from the AI agents all the way down to the systems of record. Now customers run their entire businesses on Podium. Reinventing our product wasn't even the hardest part. We had to reinvent how the entire company operated. When you've been running a company for over a decade, the operational inertia is incredibly powerful. We had to rethink how we build products, how we sell, how we serve customers, and how we run the business. We effectively started a second startup inside an already successful company. And we're still at the very beginning. Today, we have 20,000 AI agents working for our customers. They've helped generate more than $13 billion in revenue for those businesses. And we're barely scratching the surface of what AI will be able to do for them. And we've done all of this while becoming profitable and generating free cash flow. We're building Podium to be around for decades. Huge credit to our team for having the guts to make this bet and the ability to actually pull it off. We still have a long way to go and tons of problems to solve, but it's so damn fun to be winning when the opportunity ahead is this big. In the age of AI, you can either disrupt your own company or wait for somebody else to do it. I'd rather do it ourselves.
229
96
22
1,583
187,077
The good news: you can mid-train yourself to the frontier! At a fraction of the cost! Without distillation! What we did: We took Qwen3.5 14B and mid- and post-trained it to beat every other open model and perform as well as the very best frontier models. This is no simple toy task or format fixing, we are training the model to understand and reason over DNA sequences. Bio is the next frontier for AI and universal access will be crucial to make progress. The fun part: because we train the model on both natural language and DNA sequences you can actually 'chat to your DNA': Ask it questions, ponder the impact of mutations, or ask it to generate a small sequence. We'll soon see the frontier labs move away from math and cyber towards bio. It already started, and we believe researchers should have access to capable models. We are releasing the full recipe, so you can do it, too! Model: huggingface.co/HuggingFaceBi… Blog: huggingface.co/spaces/Huggin…
38
123
10
1,309
80,669