Brief

Researchers have developed a method to run large language models using fully homomorphic encryption (FHE), allowing computations on encrypted data without decryption. This breakthrough enables privacy-preserving AI applications where sensitive user data can be processed securely. The approach integrates FHE techniques with existing LLM architectures, paving the way for confidential AI services. AI

Hugging Face has released the code and weights for SD-Small and SD-Tiny, two smaller versions of its Stable Diffusion model. These models were created using knowledge distillation, a technique that trains a smaller model to mimic the behavior of a larger one. The goal is to make powerful image generation models more accessible and efficient for researchers and developers. AI

Google Research has introduced ReasoningBank, a novel framework designed to enhance AI agents' ability to learn from their experiences, both successes and failures, after deployment. This system distills generalizable reasoning strategies from past interactions, allowing agents to continuously improve and avoid repeating mistakes. Separately, new research explores optimizing multi-agent communication through latent representations and introduces Agent Evolving Learning (AEL) for agents operating in open-ended environments, focusing on how to effectively use remembered information. Additionally, DeepSeek has released preview models of its V4 series, offering large context windows and advanced capabilities at a significantly lower cost than comparable frontier models. AI

Meta AI has released Llama 2, a new large language model that is expected to significantly impact the LLM landscape. This release includes a new NeRF model called Zip-NeRF, capable of generating 3D scenes from 2D images. The hosts also discussed new functionalities from OpenAI and compared them with Anthropic's Claude 2. AI

The Hugging Face Diffusers library celebrated its first anniversary, marking a significant milestone in the open-source AI community. Since its launch, Diffusers has become a pivotal tool for researchers and developers working with diffusion models, enabling easier experimentation and deployment of generative AI applications. The library's success highlights the growing importance of accessible and collaborative platforms for advancing AI research and development. AI

Meta AI has released Llama 4, a new family of open-source large language models, available on Hugging Face. This release includes Llama Guard 4, a model specifically designed for safety, and two other models, Maverick and Scout. The availability of these models on Hugging Face Hub facilitates broader access and experimentation within the AI community. AI

Hugging Face has released an open-source model called "os-llms" designed for text generation. This model aims to foster a more collaborative and accessible ecosystem for large language models. The release emphasizes community involvement and aims to democratize access to powerful AI tools. AI

The "Practical AI" podcast discusses the recent surge in generative models, highlighting open-source advancements like Stable Diffusion XL and Zeroscope XL. Hosts Daniel and Chris predict that open models will eventually dominate the AI landscape, similar to open-source software. They also address the emerging challenges associated with this rapid progress, including cybersecurity risks, impacts on productivity, and broader cultural implications. AI

The Hugging Face Open LLM Leaderboard has updated its evaluation methodology to include the MMLU benchmark, a comprehensive test of language model knowledge across 57 subjects. This change aims to provide a more robust assessment of model capabilities by incorporating a wider range of academic and professional domains. The leaderboard now uses a weighted average of MMLU scores alongside existing benchmarks to rank open-source large language models. AI

Hugging Face has released new adapter models for their MMS (Massively Multilingual Speech) ASR system. These adapters are designed to improve performance on low-resource languages, enabling better speech recognition for a wider range of linguistic communities. The release focuses on making ASR technology more accessible and effective for languages with limited existing training data. AI

Researchers have demonstrated the effectiveness of Transformer models for time series forecasting tasks. The Autoformer architecture, specifically designed for this purpose, shows strong performance by decomposing the time series into seasonal and trend components. This approach allows for more accurate predictions by handling complex temporal dependencies. AI

Hugging Face's Open LLM Leaderboard is exploring the use of large language models (LLMs) for data labeling, aiming to replicate human-level accuracy. This approach could significantly speed up and reduce the cost of data annotation for training AI models. The blog post discusses the potential and challenges of using LLMs in this capacity, particularly in comparison to traditional human annotators. AI

Amplitude, a company known for its product analytics, is focusing heavily on integrating AI into its offerings. They are exploring methods beyond traditional Reinforcement Learning from Human Feedback (RLHF), which relies on explicit, often costly, and potentially biased user input. Instead, Amplitude advocates for learning from real user behavior within products, citing examples like GitHub Copilot and Midjourney, where implicit feedback is gathered naturally through user interaction. This approach aims to provide more authentic and cost-effective data for training AI models, potentially making AI analytics more crucial than AI itself. AI

The Falcon large language model has been integrated into the Hugging Face ecosystem. This integration makes the model more accessible to developers and researchers. Falcon is known for its strong performance on various benchmarks and its open-source nature. AI

OpenAI has developed a new method called process supervision to improve AI's mathematical reasoning capabilities. This technique rewards each correct step in a problem-solving process, rather than just the final answer, leading to better performance and reduced hallucinations. The company found that process supervision not only enhances accuracy but also offers alignment benefits by directly training models to produce human-endorsed reasoning chains. OpenAI has released its dataset to encourage further research into this promising approach. AI

The safetensors library, developed by Hugging Face in collaboration with EleutherAI and Stability AI, has undergone a security audit by Trail of Bits, confirming its safety. This audit allows the organizations to move towards making safetensors the default format for saving and loading machine learning models, replacing the less secure pickle format used by PyTorch. The library offers benefits such as faster loading times and lazy loading capabilities, and will now be installed by default in the transformers library. AI

MosaicML has released MPT-7B, an open-source transformer model trained on one trillion tokens that matches LLaMA-7B's quality and is commercially licensed. This model boasts an impressive context length of up to 84,000 tokens, significantly exceeding limitations found in models like GPT-3. MosaicML also open-sourced its LLM Foundry codebase used for training and evaluation, alongside three fine-tuned versions of MPT-7B, including one specialized for long-form storytelling. AI

Erin Mikail Staples discussed the creation of instruction-tuned Large Language Models at ODSC East. The conversation focused on the critical role of human feedback in this process. Staples also highlighted the significance of open data and practical tools for data annotation and fine-tuning custom generative AI models. AI

Hugging Face has released Q8-Chat, a new generative AI model optimized for Intel Xeon CPUs. This model aims to provide an efficient AI experience directly on standard server hardware without requiring specialized GPUs. The development focuses on making powerful AI capabilities more accessible and cost-effective for a wider range of applications. AI

Hugging Face has released a new guide detailing how to run a ChatGPT-like chatbot on a single AMD GPU using ROCm. This enables users with consumer-grade hardware to deploy powerful conversational AI models locally. The guide focuses on optimizing performance and accessibility for individuals and smaller organizations. AI

The RWKV (Receptance Weighted Key Value) project introduces a novel architecture that revives Recurrent Neural Networks (RNNs) while incorporating advantages typically found in Transformers. This approach aims to overcome the scaling limitations of traditional Transformers, particularly in training and inference, while maintaining competitive performance on reasoning benchmarks. The RWKV project is characterized by its distributed, international, and largely volunteer-driven community, drawing parallels to early EleutherAI efforts. AI

Hugging Face has released StarCoder, a new large language model specifically trained for code generation. This model is built on the StarChat architecture and has been trained on a massive dataset of permissively licensed code from GitHub. StarCoder aims to provide developers with a powerful and accessible tool for various coding tasks. AI

Replit has open-sourced its new code-focused large language model, replit-code-v1-3b. This model, which is significantly smaller than OpenAI's Codex, reportedly outperforms it on the HumanEval benchmark when fine-tuned on Replit's data. The release was discussed in an interview with Replit's Head of AI, Reza Shabani, who detailed the journey of training the model and its potential applications for developers. AI

Mark Kurtz discusses the significant advancements in optimizing large AI models for CPU inference, highlighting that a substantial portion of model parameters often do not impact outputs. This optimization work, particularly through tools like Neural Magic's SparseML and SparseGPT, enables running complex generative AI models on standard hardware, reducing the reliance on expensive GPUs and making AI more accessible. AI

Databricks has released Dolly 2.0, an instruction-following large language model that is fully open source and commercially viable. Unlike LLaMA, Dolly 2.0's license permits business use, addressing a key limitation of previous open models. The model was fine-tuned on a human-generated instruction dataset and can be customized for specific data and styles, with Databricks offering a notebook to facilitate this process for approximately $30 in 30 minutes. AI

Hugging Face has released a guide on how to run the new open-source IF (Image-to-Image) model using their diffusers library on a free tier Google Colab instance. This allows users to experiment with the model's capabilities without requiring powerful local hardware. The guide provides practical steps for setting up the environment and running inference, making advanced image generation accessible to a wider audience. AI

Hugging Face has released a new blog post detailing how to perform graph classification tasks using Transformer models. The post provides a practical guide, likely aimed at researchers and developers, on leveraging the power of Transformers for analyzing graph-structured data. This approach could open new avenues for applying advanced deep learning techniques to domains where graph data is prevalent. AI

Meta AI has released its Segment Anything Model (SAM), a significant advancement in computer vision, which includes the model, weights, data, and a demo website. This open-source release is notable for its extensive dataset, containing significantly more images and masks than previous datasets. The podcast features Joseph Nelson of Roboflow discussing SAM's capabilities, including its zero-shot transfer and promptability, and demonstrating its integration into Roboflow's platform. The discussion also touches upon the broader landscape of multimodal AI and the remaining challenges in computer vision. AI

Hugging Face has released StackLLaMA, an open-source model trained on code and natural language. This model is designed to assist developers with coding tasks, offering capabilities such as code generation and explanation. The release aims to provide a powerful, accessible tool for the AI development community. AI

Researchers have demonstrated a method for training and analyzing language models using Reinforcement Learning from Human Feedback (RLHF). The process involves using the TRLX library for RLHF fine-tuning and TransformerLens for mechanistic interpretability. This approach was used to fine-tune a GPT-2 model to generate negatively biased movie reviews and then analyze the model to identify specific network regions responsible for this behavior. AI

Hugging Face has released a new guide detailing how to achieve fast inference for large language models like BLOOMZ using Habana Gaudi2 accelerators. The guide provides practical steps and optimizations for developers looking to leverage this hardware for efficient LLM deployment. This collaboration aims to make powerful AI models more accessible and performant on specialized hardware. AI

The Alignment Research Center's (ARC) evaluation team has officially spun off to form a new independent nonprofit organization named METR (Model Evaluation & Threat Research). METR will continue its work on evaluating frontier AI systems, focusing on their autonomous capabilities and potential threats, including AI self-improvement and evasion of oversight. The organization, led by Beth Barnes, has previously partnered with leading AI labs like OpenAI and Anthropic for evaluations and aims to develop rigorous testing methodologies to ensure AI safety before widespread deployment. AI

Iceland has partnered with OpenAI to leverage GPT-4 for the preservation of the Icelandic language, which is at risk of decline due to digitalization. A team of 40 volunteers is using Reinforcement Learning from Human Feedback (RLHF) to train GPT-4 on proper Icelandic grammar and cultural nuances. This initiative aims not only to safeguard Icelandic but also to create a model for preserving other low-resource languages globally, preventing an "AI divide." AI