Raphaël Gazzotti

🚨Breaking: Someone just open sourced a knowledge graph engine for your codebase and it's terrifying how good it is. It's called GitNexus. And it's not a documentation tool. It's a full code intelligence layer that maps every dependency, call chain, and execution flow in your repo -- then plugs directly into Claude Code, Cursor, and Windsurf via MCP. Here's what this thing does autonomously: → Indexes your entire codebase into a graph with Tree-sitter AST parsing → Maps every function call, import, class inheritance, and interface → Groups related code into functional clusters with cohesion scores → Traces execution flows from entry points through full call chains → Runs blast radius analysis before you change a single line → Detects which processes break when you touch a specific function → Renames symbols across 5+ files in one coordinated operation → Generates a full codebase wiki from the knowledge graph automatically Here's the wildest part: Your AI agent edits UserService.validate(). It doesn't know 47 functions depend on its return type. Breaking changes ship. GitNexus pre-computes the entire dependency structure at index time -- so when Claude Code asks "what depends on this?", it gets a complete answer in 1 query instead of 10. Smaller models get full architectural clarity. Even GPT-4o-mini stops breaking call chains. One command to set it up: `npx gitnexus analyze` That's it. MCP registers automatically. Claude Code hooks install themselves. Your AI agent has been coding blind. This fixes that. 9.4K GitHub stars. 1.2K forks. Already trending. 100% Open Source. (Link in the comments)

Today, we’re introducing Pomelli’s latest feature update, ‘Photoshoot’ With Photoshoot, you can start from a single image of your product and easily create high quality, customized product shots to elevate your marketing. Available free of charge in the US, Canada, Australia & New Zealand! Get started with Pomelli today at labs.google/pomelli

This aired tonight to 1 billion people in China. A year ago these robots could barely wave a handkerchief, now they can do backflips and kung fu with nunchucks. Physical intelligence is the next frontier.

The poster session at the #EBRAINSSummit2025 is happening now! Join us in the exhibition hall to discover posters about data curation, EBRAINS National Nodes, brain atlases, and more!

🏆 At the #EBRAINSSummit2025 we celebrate the work of researchers across the EBRAINS network with the Best Abstract Awards! The 3 winners are: 🔬 For neuroscientific & medical resources: 
• Layer-specific cell counts in BigBrain, Sebastian Bludau & Timo Dicksheid @fz_juelich 🤖 For Technology & AI:
• Bids2ebrains, Renqing Cuomao @EPFL_en 
• The Virtual Aging Brain, Amirhossein Esmaeili @uniamu

Cara grew from 40k to 650k users in a week because artists are fed up with Meta's AI policies We're 700k now! - techcrunch.com/2024/06/06/a-s… 1/

Introducing Sora, our text-to-video model. Sora can create videos of up to 60 seconds featuring highly detailed scenes, complex camera motion, and multiple characters with vibrant emotions. openai.com/sora Prompt: “Beautiful, snowy Tokyo city is bustling. The camera moves through the bustling city street, following several people enjoying the beautiful snowy weather and shopping at nearby stalls. Gorgeous sakura petals are flying through the wind along with snowflakes.”

Google has already started forcing Web Integrity into Chromium despite it being a 'proposal' With WEI, users can be denied access for using non-approved browsers or hardware The open Internet is officially dead as soon as this is commonly implemented

I can't believe I've just fine-tuned a 33B-parameter LLM on Google Colab in a few hours.😱 Insane announcement for any of you using open-source LLMs on normal GPUs! 🤯 A new paper has been released, QLoRA, which is nothing short of game-changing for the ability to train and fine-tune LLMs on consumers' GPUs. In a few words: QLoRA reduces the memory usage of LLM fine-tuning without any performance tradeoffs compared to standard 16-bit model fine-tuning. This method enables 33B model fine-tuning on a single 24GB GPU and 65B model fine-tuning on a single 46GB GPU. This is incredible! 😍 More specifically, QLoRA uses 4-bit quantization to compress a pre-trained language model. The LM parameters are then frozen, and a relatively small number of trainable parameters are added to the model in the form of Low-Rank Adapters. During finetuning, QLoRA backpropagates gradients through the frozen 4-bit quantized pretrained language model into the Low-Rank Adapters. The LoRA layers are the only parameters being updated during training. Read more about LoRA in the original LoRA paper (arxiv.org/abs/2106.09685). 🤓 QLoRA has one storage data type (usually 4-bit NormalFloat) for the base model weights and a computation data type (16-bit BrainFloat) used to perform computations. QLoRA dequantizes weights from the storage data type to the computation data type to perform the forward and backward passes, but only computes weight gradients for the LoRA parameters, which use 16-bit bfloat. The weights are decompressed only when they are needed, therefore the memory usage stays low during training and inference. Beautiful!😱 QLoRA tuning is shown to match 16-bit finetuning methods in a wide range of experiments. In addition, the Guanaco models, which use QLoRA finetuning for LLaMA models on the OpenAssistant dataset (OASST1), are state-of-the-art chatbot systems and are close to ChatGPT on the Vicuna benchmark. This is an additional demonstration of the power of QLoRA tuning. Their Guanaco models are reaching 99.3% of the performance level of ChatGPT while only requiring 24 hours of fine-tuning on a single GPU. You can actually do it in Google Colab. 📚 Links- QLoRA Paper - arxiv.org/pdf/2305.14314… Colab for inference - colab.research.google.com/drive/1ge2F1QS… Colab for fine-tuning - colab.research.google.com/drive/1VoYNfYD… GitHub Repository- github.com/artidoro/qlora Use it with HuggingFace - huggingface.co/blog/4bit-tran…

We integrated ChatGPT with our robots. We had a ton of fun building this! Read on for the details:

She is not real. Generative AI is mindblowing, but there's a catch: Anything serious will cost you an arm and a leg. But there's a solution that will save your wallet:

Please RT We're hiring a student researcher here at Google Research. If you are a PhD student interested in language modeling and have a strong background in stats, feel free to reach out via email. careers.google.com/jobs/results/1…

Want to use #LOD in #DH, but can't decide on a tool? Check out v.1.0.0 of our tool comparison chart: zenodo.org/record/7738424 Thanks to my @nfdi4culture colleague Robert Nasarek for being our in-person presenter this morning @DHdKonferenz & putting together the video highlights:

/1 How Discord Stores Trillions Of Messages The diagram below shows the evolution of message storage at Discord: MongoDB ➡️ Cassandar ➡️ ScyllaDB

On this day in 1989: Tim Berners-Lee shares the first proposal for World Wide Web w3.org/History/1989/p… #InternetHistory

A search engine that scrapes Sci-Hub, Library Genesis, and other shadow libraries. Hoist the colors! 🏴‍☠️📚 annas-archive.org

@wimmics publishes in-house guidelines for deploying @OpenLink #VirtuosoRDBMS OS #triplestore with Docker, Apache front-end, metadata... Based on exprnc from 5 projects including #DBpedia FR. t.ly/MNrBq @inria_sophia @Univ_CotedAzur @Laboratoire_I3S @cnrs @kidehen

Before we reach Human-Level AI (HLAI), we will have to reach Cat-Level & Dog-Level AI. We are nowhere near that. We are still missing something big. LLM's linguistic abilities notwithstanding. A house cat has way more common sense and understanding of the world than any LLM.

To develop its #NLP activity the @wimmics team at @inria_sophia @Univ_CotedAzur @Laboratoire_I3S looks for candidates for permanent researcher and assistant professor positions in NLP and argumentation in the beautiful French Riviera. Interested? Contact @ECabrio @serena_villata

@SW_Journal Is there any news about this maintenance? It's been several days, and error messages are also popping up.