✷ The archive · 11 labs · 3,432 dispatches
The labs
No more opening a dozen official blogs every morning. First-hand releases from OpenAI, Anthropic, DeepMind and the rest, each with its substance pulled out.
Extracting Concepts from GPT-4
OpenAI has developed scalable sparse autoencoder methods to decompose GPT-4's internal representations into 16 million human-interpretable features.
Anthropic Election Integrity Testing and Mitigation Framework
Anthropic has implemented a multi-layered testing and mitigation process combining expert qualitative analysis and automated evaluations to safeguard election integrity in its AI models.
Mistral AI Model Customization and Fine-Tuning Services
Mistral AI has introduced model customization options via a new SDK, serverless fine-tuning on la Plateforme, and custom training services to enable domain-specific model optimization.
Mistral AI Fine-tuning Hackathon Announcement
Mistral AI announced a virtual fine-tuning hackathon running June 5‑30 2024, offering €2,500 in API credits to the top three projects that use its new fine‑tuning API.
Hugging Face NPC-Playground: Integrating LLM-Powered NPCs in 3D Environments
Hugging Face introduces NPC-Playground, a 3D demo utilizing Cubzh and Gigax to enable realistic, LLM-driven NPC interactions through function calling and Lua scripting.
Anthropic Launches Claude in Canada
Anthropic has expanded the availability of Claude, including the Claude 3 model family and API, to users and businesses in Canada.
Intel Gaudi Assisted Generation Support
Hugging Face has integrated assisted generation (speculative sampling) into Optimum Habana, enabling up to 2x speedups for large transformer-based models on Intel Gaudi processors.
Hugging Face Spaces Secrets Security Update
Hugging Face disclosed a security incident involving unauthorized access to a subset of Spaces secrets, leading to the revocation of affected tokens and a transition to fine-grained access tokens.
Ollama and Continue: Setting Up an Open-Source AI Code Assistant
Ollama and Continue enable developers to build a fully open-source, local AI coding assistant within VS Code and JetBrains using models like Codestral, Llama 3, and DeepSeek Coder.
OpenAI Disrupts Covert AI Influence Operations
OpenAI terminated accounts linked to five covert influence operations that used AI to generate deceptive content, though none achieved significant audience reach.
OpenAI Introduces ChatGPT Edu for Universities
OpenAI has launched ChatGPT Edu, a version of ChatGPT designed for universities to deploy AI at scale with enterprise-grade security, data privacy, and access to GPT-4o.
OpenAI for Nonprofits Initiative
OpenAI has launched OpenAI for Nonprofits, providing discounted access to ChatGPT Business and Enterprise to help nonprofit organizations overcome operational and funding constraints.
MavenAGI Automated Customer Support Agents powered by OpenAI
MavenAGI has launched AI customer service agents built on GPT-4 that automate up to 93% of support questions and reduce the average cost per ticket from $40 to $8.
Codestral Release Notes
Mistral AI has released Codestral, a 22B open-weight generative AI model specifically designed for code generation across 80+ programming languages.
OpenAI and WAN-IFRA Launch The Newsroom AI Catalyst Program
OpenAI and WAN-IFRA have launched The Newsroom AI Catalyst, an accelerator program providing funding and technical assistance to 128 newsrooms globally to fast-track AI adoption and improve journalistic sustainability.
Mistral AI Non-Production License (MNPL) Announcement
Mistral AI has introduced the Mistral AI Non-Production License (MNPL) to allow non-commercial and research use of its technology while ensuring sustainable commercial partnerships for those building businesses on its models.
OpenAI and The Atlantic Partnership for News Integration
OpenAI and The Atlantic have entered a strategic partnership to integrate The Atlantic's premium news content into ChatGPT and other OpenAI products, featuring attribution and links to full articles.
OpenAI and Vox Media Strategic Partnership
OpenAI and Vox Media have entered a strategic partnership to integrate Vox Media's journalistic content into ChatGPT and collaborate on AI-driven products for consumers and advertisers.
Benchmarking Text Generation Inference
Hugging Face introduces a benchmarking tool for Text Generation Inference (TGI) to help developers profile throughput and latency trade-offs to optimize LLM deployment costs and performance.
Anthropic Appoints Jay Kreps to Board of Directors
Anthropic has appointed Jay Kreps, co-founder and CEO of Confluent, to its Board of Directors to support the company's enterprise growth and data infrastructure scaling.
OpenAI Board Forms Safety and Security Committee
OpenAI has established a Safety and Security Committee to oversee critical safety and security decisions for its projects and operations as it begins training its next frontier model.
Training and Finetuning Embedding Models with Sentence Transformers
Hugging Face provides a comprehensive guide on using the Sentence Transformers library to train and finetune embedding models for tasks like semantic search and RAG, introducing the new SentenceTransformerTrainer.
xAI Series B Funding Round
xAI has raised $6 billion in Series B funding to accelerate the development of advanced AI systems and build necessary infrastructure to bring its products to market.
Falcon 2 11B Release Notes
TII has released Falcon 2 11B, a series of smaller, high-performance models including a pretrained LLM and a Vision-Language Model (VLM) trained on over 5,000 billion tokens.
CyberSecEval 2: Evaluating Cybersecurity Risks in Large Language Models
Hugging Face and Meta have introduced CyberSecEval 2, a comprehensive framework to evaluate LLM susceptibility to code interpreter abuse, offensive cybersecurity capabilities, and prompt injection attacks.
OpenAI and News Corp Multi-Year Global Partnership
OpenAI and News Corp have entered a multi-year agreement allowing OpenAI to integrate News Corp's current and archived content from major publications like The Wall Street Journal and The Times into its products to improve information reliability.
Hugging Face AWS Inferentia2 Integration
Hugging Face has expanded support for AWS Inferentia2, enabling the deployment of over 100,000 models via Amazon SageMaker and introducing new Inf2 instance options for Hugging Face Inference Endpoints.
OpenAI Safety Practices
OpenAI has detailed ten core safety practices used to develop and deploy frontier AI models, emphasizing a scientific approach that integrates safety into every stage of the model lifecycle.
Hugging Face and Microsoft Strategic Collaboration Update May 2024
Hugging Face and Microsoft have expanded their strategic partnership to integrate popular open models into the Azure Model Catalog, enable AMD MI300X GPU support on Azure, and introduce a VS Code integration for Hugging Face Spaces.
Hugging Face Integration for AMD Instinct MI300 GPU
Hugging Face has enabled first-class integration for the AMD Instinct MI300 GPU, delivering up to 3x faster inference and 2x faster fine-tuning for Llama 3 70B compared to the MI250.
Dell Enterprise Hub: On-Premise AI Deployment and Training
Hugging Face and Dell have launched the Dell Enterprise Hub, a platform that simplifies the on-premise training and deployment of open-source large language models on Dell hardware.
Hugging Face Spaces Dev Mode Release
Hugging Face has introduced Spaces Dev Mode, allowing PRO subscribers to connect via VS Code or SSH for real-time code editing without requiring git pushes for every change.
Anthropic maps internal concepts of Claude 3 Sonnet using large-scale dictionary learning
Anthropic disclosed that it extracted millions of interpretable features from Claude 3 Sonnet, revealing how concepts are represented inside a production‑grade LLM and showing that manipulating these features can change model behavior, a step toward safer AI.
Krishna Rao joins Anthropic as Chief Financial Officer
Anthropic has appointed Krishna Rao as Chief Financial Officer to lead financial strategy and operations during a period of enterprise growth and international expansion.
Anthropic Responsible Scaling Policy Reflections
Anthropic provides an operational update on its Responsible Scaling Policy (RSP), detailing the implementation of AI Safety Levels (ASL) and the frameworks used to mitigate catastrophic risks in frontier models.
Firebase Genkit with Ollama Support
Google has released Firebase Genkit, an open-source framework for building AI-powered apps, featuring native support for Ollama to run Google's Gemma model locally.
OpenAI ChatGPT Voice Selection Process
OpenAI detailed the five-month casting process used to select the five professional voice actors for ChatGPT, while clarifying that the voice 'Sky' was not intended to mimic Scarlett Johansson.
ChatGPT Data Analysis Improvements May 2024
OpenAI has introduced enhancements to data analysis in ChatGPT, including direct cloud storage integration, interactive tables, and customizable charts, powered by GPT-4o.
OpenAI and Reddit Partnership
OpenAI and Reddit have partnered to integrate real-time Reddit content into ChatGPT and implement OpenAI's AI models to enhance Reddit's user and moderator features.
Hugging Face KV Cache Quantization
Hugging Face has introduced KV Cache Quantization in the Transformers library to reduce memory usage for long-context text generation with minimal impact on model quality.
Canva Magic Studio Integration with OpenAI
Canva has integrated OpenAI's GPT-4 and API to power its Magic Studio AI suite, which has been used over 5 billion times to simplify design for non-professional creators.
Mike Krieger Joins Anthropic as Chief Product Officer
Anthropic has appointed Mike Krieger, co-founder and former CTO of Instagram, as Chief Product Officer to lead product engineering, management, and design.
OpenAI Leadership Change: Ilya Sutskever Departs, Jakub Pachocki Named Chief Scientist
OpenAI has announced that co-founder Ilya Sutskever is leaving the company, with Jakub Pachocki appointed as the new Chief Scientist to lead research efforts toward AGI.
Introducing the Open Arabic LLM Leaderboard
Hugging Face and the Technology Innovation Institute have launched the Open Arabic LLM Leaderboard (OALL) to provide a specialized, open evaluation platform for Arabic Large Language Models.
PaliGemma Vision Language Model Release
Google has released PaliGemma, a family of open vision-language models combining SigLIP-So400m and Gemma-2B to enable tasks like image captioning, visual question answering, and object detection.
Hugging Face and LangChain Launch langchain_huggingface Partner Package
Hugging Face and LangChain have released langchain_huggingface, a jointly maintained partner package to streamline the integration of Hugging Face models into the LangChain ecosystem.
Claude now available in the EU – Anthropic expands access to its AI assistant
Anthropic announced that Claude, its AI assistant, is now available to individuals and businesses across Europe via Claude.ai, an iOS app, and a Team plan, extending the Claude 3 model family to the EU market.
OpenAI GPT-4o Release Notes
OpenAI has released GPT-4o, an omni-model trained end-to-end across text, vision, and audio to enable natural human-computer interaction with human-like response latencies.
OpenAI Spring Update: Introducing GPT-4o
OpenAI has released GPT-4o, a new flagship model capable of real-time reasoning across audio, vision, and text, while expanding advanced capabilities to ChatGPT free users.
OpenAI GPT-4o Release and ChatGPT Free Tier Enhancements
OpenAI has launched GPT-4o, a flagship model providing GPT-4 level intelligence with improved speed and multimodal capabilities across text, voice, and vision, now available to both free and paid users.