Full-text search across article titles and summaries. Refine with sector, year, content type, and reading-time filters.
The Federal Trade Commission opened an inquiry into the investments of Big Tech companies that provide cloud services to smaller AI companies like OpenAI and Anthropic.
Google DeepMind has developed an AI model called AMIE (Articulate Medical Intelligence Explorer) to assist in healthcare. Trained on medical texts and dialogues, AMIE will provide explanations of medical conditions and potentially aid in diagnostics. AMIE outperformed human physicians in a controlled trial. Its effectiveness in real-world scenarios remains to be tested.
Another multimodal model trained on Phi-2 and SigLIP. This one achieves extremely promising performance and is small enough to run on-device.
OpenAI's new embeddings can change size based on your needs. This is theorized to be due to the nesting doll method of learning, which learns features at various granularity.
Vivim is a new framework for medical video object segmentation that effectively processes long video sequences. Vivim compresses spatiotemporal data efficiently utilizing the state space model approach, delivering faster and more accurate segmentation results compared to traditional methods.
WebDataset is a format that combines multiple records together. Anything with the same prefix is treated as an instance. It is ideal for streaming and fast data loading.
An incomplete but promising set of tasks for learning how to build language models from scratch.
This project proposes a novel approach for enhancing transformers that uses unrelated data from different modalities, such as improving an image model using audio data. Multimodal Pathway uniquely connects transformers of two different modalities, allowing a target modality to benefit from the strengths of another.
ColBERT is an excellent model for embedding queries and index data for RAG applications. This post explores the intuition behind the method and performs some benchmarks.
Pix2Gestalt is a framework designed for zero-shot amodal segmentation. It is adept at reconstructing complete object shapes and appearances when they are partially obscured. Leveraging large-scale diffusion models, Pix2Gestalt excels in challenging scenarios, including artistic images that defy usual patterns.
Many groups are trying to grapple with the challenges that come with rapidly advancing AI technology. Some of their moves can be seen as anti-competitive. It is a complex problem with valid arguments on both sides.
Sponsored
Is your enterprise data trapped in hard-to-use formats? This tool makes it effortlessly available in every major vector database and
Tachyum is set to start mass production of its Prodigy Universal Processor, a versatile unit that combines CPU, GPU, and TPU capabilities and promises to significantly boost performance in cloud, high-performance computing, and AI applications. It could potentially disrupt the AI market.
Turn your ideas into visually compelling branded social posts.
Inspired by efforts on the Llama models, this project uses pure Rust to run inference for Mamba on the CPU.
Hugging Face has partnered with Google Cloud to give easy access to models and compute for customers of both platforms.
OpenAI released a new embeddings model, a better and cheaper GPT-3.5 Turbo, and a less lazy GPT-4 Turbo preview. It has improved pricing and tried to improve the GPT-4 Turbo preview model to prevent the laziness issue in code generation.
The vector database company that powers portions of ChatGPT and X's “More like this” has raised money to improve its Rust-based vector store and expand its enterprise offerings.
Soft robotics is a challenge due to the large control space. This work introduces a simulator with various tasks for manipulating soft “ditto” like objects. It contains utilities, visualization, and some strong baselines.
Researchers have developed a new method for creating scene graphs, making the process faster and more efficient. Their transformer-based technique focuses on improving the way the model understands and connects different elements in an image, leading to better performance on challenging tasks.
SpeechGPT-Gen is a new speech synthesis model that efficiently handles the complexities of language and voice characteristics.
Sponsored
Microsoft is marketing its Fabric Lakehouse architecture as an ideal way to manage Power BI semantic models. This tech talk [12] will examine these claims in light of AtScale’s TPC-DS benchmark [13], measuring Power BI performance on Microsoft Fabric on 100GB, 1TB, and
LLM Steer is a library for controlling language model generation with sentiment vectors. This means you can embed a concept like “thoughtful” and make the generation more thoughtful. It tends to work better than prompting in general, but more study is needed.
RoMa is a library for PyTorch for dealing with tricky rotations. It is reasonably efficient and can be used to help with any project that uses 3D data.
AgentBoard is a benchmark designed for multi-turn LLM agents, complemented by an analytical evaluation board for detailed model assessment beyond final success rates.
A walkthrough with code on writing a sparse Mixture of Experts model from scratch in Pytorch. There are diagrams and explanations for each step. The explanation of top-k routing is particularly insightful.
A survey of experts on how early adopters are using AI today and where AI is going in 2024.
Image similarity systems give a score based on how similar two images are. This work improves upon previous methods, primarily by relying on synthetic data and human preferences.
SegMamba is a model designed for 3D medical image segmentation. It offers a more efficient alternative to the Transformer architecture.
Track keywords across social media to convert leads with AI-generated auto-replies.
Researchers have developed a Shared Feature Calibration (SFC) method to enhance semantic segmentation.
Sponsored
For LegalTech Companies: Elevate your platform instantly with CaseMark’s Workflow API [7]. Seamlessly integrate our solution to offer AI-driven deposition, case, trial, and hearing summaries.
Embeddings serve as the foundation for many AI use cases. They can be expensive and slow due to rate limits. This blog walks through using Modal to scale up the embedding process efficiently and correctly.
Google, the Weizmann Institute of Science, and Tel Aviv University have unveiled Lumiere, a space-time diffusion model for realistic video generation that promises more coherent motion and better quality compared to existing AI video models. It's not yet available for public testing.
Neel Nanda is one of the pioneers of Mechanistic Interpretability (MI). This is his guide to getting into the field. It features 200 concrete open questions. MI is the study of the quantitative values of language models (e.g., actually looking at neurons). It is accessible as it doesn’t require significant computation, and is extremely promising as a research direction, although there has been minimal success to date.
Adept scaled up its Fuyu architecture and trained a larger version that is currently the 3rd best model on MMMU. The model also performs well on other tasks (e.g., achieving 40+ on HumanEval coding benchmark). There were fewer challenges with scaling this architecture than alternatives, in part due to its simplicity. However, there were still challenges with data i/o, memory bandwidth, and custom kernel development.
HEDNet is a new encoder-decoder network designed to enhance 3D object detection in autonomous vehicles, specifically addressing the challenge of sparse point distribution in 3D scenes.
The Song Describer dataset features over a thousand human-written descriptions of music recordings. This new dataset aids in evaluating music-and-language models through tasks like music captioning and text-to-music generation.
This project presents a novel prompt pool method for dialog state tracking that adapts to new user requirements without needing task identities during testing.
DiffMoog is a new audio technology that combines the versatility of a modular synthesizer with the power of neural networks. It allows for automated sound matching and custom sound creation, making it a valuable tool for audio synthesis.
A great little utility that allows you to use dictionaries with tensors in PyTorch. It improves readability and usability of code while maintaining performance.
The Turing Test and Searle's Chinese Room Thought Experiment are often cited in the discussion of AI consciousness, with the former testing if an AI's behavior is indistinguishable from a human's and the latter arguing that external behavior alone cannot prove consciousness. This debate highlights the difficulty in defining and recognizing consciousness in AI, as our understanding is primarily based on human experiences and functionalist approaches.
This article discusses evaluation metrics that don't rely on ground truth data. It explores general-purpose as well as RAG, chatbot, and summarization-specific evaluations. All metrics come with a Python implementation.
Apple is intensifying its focus on AI for the iPhone, acquiring numerous AI companies, and increasing AI job postings.
BreezeML is an AI governance platform that enables compliance teams to specify and continually monitor governance policies over every AI workflow in its organization without relying on manual and tedious coordination with data science teams.
Researchers have explored asynchronous Local-SGD (a variant of federated averaging) for training language models.