Full-text search across article titles and summaries. Refine with sector, year, content type, and reading-time filters.
This paper introduces a benchmark and two datasets for scene and handwritten texts. Featuring original, corrupted, and assistant images, the novel Global Structure-guided Diffusion Model (GSDM) leverages text structure to efficiently recover clean texts. It shows significant improvements in recognition accuracy and image quality.
Norton is a new approach to improving long-term video understanding by addressing the multi-granularity noisy correspondence issue in video-language studies.
MF-MOS is a new model for better detecting moving objects in autonomous driving using LiDAR technology. It uniquely separates motion and semantic information, using residual maps for motion capture and range images for semantic guidance.
Monte Carlo Tree Search (MCTS) is a state-of-the-art search algorithm. It was used in AlphaGo and researchers are now working to integrate it with language models. MCTS is notoriously tricky to implement.
AI has gotten good enough to accomplish multiple tasks in under a minute, which means organizations and employees are going to have to emphasize using AI for good, not bad.
FireLlava is a new open vision model trained on data that makes it possible to use for commercial tasks. It matches the performance of the original Llava, but not yet to the level of Llava 1.5.
ServiceNow, a $150b market cap company, last week announced that its gen AI products drove the largest net new ACV contribution for its first full quarter of any of its new product family releases ever, including its original Pro SKU. We’re already seeing enterprise-level AI applications drive meaningful revenue growth – that’s cool.
Sponsored
Fly to Hawaii for $149, Greece for $289, or anywhere else. Get 93%
before midnight. Comes with 3-day money-back guarantee*
A library that allows for creating autonomous AI agents in Python.
GPAvatar introduces a new method for creating 3D head avatars from images that use a dynamic point-based expression field and a Multi Tri-planes Attention module.
Automate resume screening with AI.
A delightful gadget that shows the time via a poem has been launched. It uses clever prompt caching and is beautifully designed.
Apple is preparing major artificial intelligence features for iOS 18. It is using OpenAI's ChatGPT API for internal testing to develop its own AI models. The first beta of iOS 17.4 revealed a new SiriSummarization framework that integrates AI into the Messages app for functionalities like summarization and auto-completion.
Opera announced the upcoming launch of its AI-powered Opera One browser for iOS in Europe, following Apple's decision to allow non-WebKit-based browsers on iOS due to the European Digital Markets Act.
A challenge in training language models is finding a diverse enough dataset for the task. Visualizing this data is even harder. This cool tool uses fast clustering and topic modeling to enable the exploration of data to improve filtering and overall quality.
“Up captioning” is an idea used in many AI systems to improve labels while training. This work from Apple applies this to pre-training by rephrasing C4 as Q/A pairs, instructions, and more. The study found the rephrasing step sped up convergence by 10x, making the model dramatically more sample efficient - at the cost of the rephrasing step.
This paper offers a detailed survey of the latest advancements in continual learning, which focuses on continuously adapting to new data while retaining previous knowledge.
Sponsored
An up-to-date backlog helps your team prioritize the right projects, but managing it is a ton of work — which shouldn’t fall on you. Let Height handle the project legwork of backlog refining, bug triage, and chasing statuses.
Training models with lower precision is faster and cheaper, but unstable. There has been lots of great recent research on quantized training. This repository builds on those foundations with easy-to-read and hackable code to enable float8 training.
DeFlow is a new method for estimating 3D motion in scenes, a skill crucial for autonomous driving. It overcomes the limitations of traditional voxelization by transitioning from voxel-based to point features using a Gated Recurrent Unit refinement and introducing a novel loss function.
This project introduces MAcro Recommendation Graph (MAG) and Macro Graph Neural Networks (MacGNN). These approachs tackle the computational complexity of Graph Neural Networks by grouping similar behavior patterns into macro nodes, significantly reducing node count.
Enchanted is an open source app for iOS that allows you to run language models on your mobile device.
SliceGPT introduces a novel post-training sparsification scheme to reduce the resource demands of large language models. By replacing weight matrices with smaller ones and reducing embedding dimensions, it can remove up to 25% of model parameters in major models like LLAMA2-70B and OPT 66B, while retaining up to 99% task performance.
OpenAI CEO Sam Altman's conflicting messages about AI's potential and limitations, including alternating between hyping up advancements like GPT-5 and tempering expectations, reflect a strategic balance between generating excitement and managing public sentiment but raise questions about the true progress and impact of AI technology.
Sponsored
Crypto accounting can get messy. Entendre Finance’s enterprise-grade AI keeps all your digital assets accounted for — automatically. Join the waitlist to try the new Onboarding Concierge!
Use an LLM directly anywhere you can type.
OpenAI is enhancing ChatGPT by allowing paid users to incorporate various GPTs into conversations, each understanding the full context of the conversation they are inserted into.
Your AI product growth consultant.
Meta's new models have been fine-tuned and instruction-tuned on significantly more code. They achieve impressive performance on HumanEval, approaching GPT-4 in some respects. The models are available under the Meta license and on HuggingFace
RWKV is an architecture alternative to the dominant Transformer model. It uses a novel linear attention mechanism that makes it extremely efficient. This new checkpoint and configuration was trained on 1T tokens and surpases many other powerful 7B Transformer based models.
OpenAI is suspected of violating European Union privacy laws. Italy's data protection authority has given it 30 days to respond to allegations following an investigation into ChatGPT.
One key challenge with 3D object generation is the diversity of items that can be created. This work uses a modified architecture to improve sample efficiency and scales up the system to work with a much larger set of objects in each 3D category.
Another potential distillation work. Importantly, this one can work on models as small as Phi-2. This means you can remove 90% of the rows and columns of weight matrices with minimal reduction to quality at almost all scales.
Meta Learning is the process of training systems that learn to learn and can quickly adapt to new tasks. This work from Google improves Meta Learning with synthetic data generated from a Universal Turing Machine and analyzes the results experimentally and theoretically.
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 [12], measuring Power BI performance on Microsoft Fabric on 100GB, 1TB and
CreativeSynth is a method for artistic image editing by seamlessly integrating text and image inputs. Its diffusion model, equipped with customized attention mechanisms, enables precise style and content manipulation while preserving the original artwork's essence.
Inconsistency Masks (IM) are a new method for image segmentation that works even with limited data. This approach, tested on the ISIC 2018 dataset, beats traditional techniques and even outperforms models trained on fully labeled datasets.
A new image distortion technique makes facial images unrecognizable to humans but still identifiable by AI, striking a balance between privacy and accuracy in biometric systems.
Researchers have improved the ability of generative models like GANs and diffusion models to create realistic hand images by adding three extra channels to training images for hand annotations.
Looking for technical sales specialist interested in AI/LLMs with 3+ years of sales experience in cloud computing/infrastructure to help drive adoption of our NVIDIA GPU cloud. Apply here [18].
Shortwave released an AI-powered autocomplete feature that's similar to Github Copilot but designed for writing emails. It learns from sent emails and provides contextual suggestions.
An open source platform for AI-native application development.
The enterprise data toolkit acquired Claypot's real-time AI tools and team for an undisclosed sum. Voltron aims to continue bringing access to data to the enterprise.