Search the archive
Full-text search across article titles and summaries. Refine with sector, year, content type, and reading-time filters.
Results
14,500 matchesMORE DATA IS BETTER, EVEN FOR ROBOTS
HOW DOES RLHF CHANGE THE INNER-WORKINGS OF TRANSFORMERS?
ANOTHER LLAMA MODEL JOINS THE ZOO
Similar to Alpaca from Stanford, Koala was trained on dialog and instruction data. However, they didn’t use distillation from the OpenAI models, but sourced their data from freely available sources on the web. Their model actually seems comparable to strong closed source models and the code to fine-tune along with the dataset are available for further study.
FOLLOW YOUR POSE: TEXT-TO-VIDEO GENERATION USING POSE-FREE VIDEOS
This paper proposes a two-stage training scheme for generating text-editable and pose-controllable character videos, using easily obtainable datasets and a pre-trained text-to-image model. The method successfully generates pose-controllable character videos and retains the editing abilities of the pre-trained model.
FAST REAL-TIME VIDEO ENHANCEMENT
ReBotNet is a video model that is designed to enhance live streams and video conferences in real-time. They use a dual-branch network that learns different types of video abstractions like location and time based ones. They then use a common decoder network to output the enhanced frames. It runs at 50 fps, uses about 5gb of memory and performs on par with models 8x larger.
BABYAGI
Babyagi is a Python AI-powered task management system.
AI IN 14 CHARTS
TAKEAWAYS FROM STANFORD’S 386-PAGE AI REPORT
BIDEN MEETING WITH ADVISORS TO DISCUSS AI
AMAZON LAUNCHES STARTUP ACCELERATOR FOR GENERATIVE AI COMPANIES
EXPEDIA INTEGRATES GPT-4
Sponsored
COMPLIANCE DOESN’T HAVE TO BE SLOW AND TEDIOUS
APPLE CEO TIM COOK TEASES AR/VR HEADSET AND MORE IN NEW INTERVIEW
NOT
A DECADE LATER, THIS VR TREADMILL IS FINALLY READY TO SHIP
HOCUS
Hocus allows users to launch disposable development environments on their own servers in a matter of seconds. Users can define their dev environments as code and launch them immediately from their browsers. Hocus serves as a capable self-hosted option for Gitpod and GitHub Codespaces and integrates with any Git provider that supports the SSH protocol.
CSS TEXT-WRAP : BALANCE : CHROME DEVELOPERS
Sponsored
APPLICATIONS WORK WITHOUT CODE CHANGES!
HOW TO DO HARD THINGS
NETWORK
SHIBA INU
FASTCHAT
'MULTI-YEAR' SAVINGS
HOW TO RUN A SHADOW LIBRARY: OPERATIONS AT ANNA’S ARCHIVE
BY OPENAI
LAGO
Lago is a usage-based billing API for product-led Saas.
FIFA DEBUTS AI-POWERED SOCCER GAME
CLEARVIEW AI GAVE 30 BILLION FACEBOOK IMAGES TO THE POLICE
Clearview AI built it’s facial recognition database using 30 billion images from Facebook. Because this database has been used by the police nearly a million times now, Clearview is now being accused of putting people in a perpetual police line-up.
AI DOESN’T HALLUCINATE, IT MAKES THINGS UP
VIDEOMAE V2: SCALING VIDEO MASKED AUTOENCODERS WITH DUAL MASKING
VideoMAE pioneers a scalable and versatile self-supervised pre-training approach for constructing video foundation models using a dual masking strategy and progressive training paradigm. Achieving state-of-the-art performance on multiple datasets, VideoMAE proves its efficacy as a general video representation learner for a wide range of downstream tasks.
LANGUAGE MODELS CAN CRITIQUE THEMSELVES?
LANGUAGE MODELS ARE THE PUPPET MASTERS
People are exploring how language models can use tools - Hugging Face has thousands of AI models hosted on their website, what if each of those was a potential tool? It turns out, you can use ChatGPT as the controller and each Hugging Face model as a specialized tool. Do you want to do a visual question answering task followed by summarization? This work is a first step towards potentially automating that task.
INCEPTIONNEXT: WHEN INCEPTION MEETS CONVNEXT
InceptionNeXt introduces a novel approach to enhance large-kernel-based CNN models' efficiency while maintaining performance by decomposing large-kernel depthwise convolutions into four parallel branches along the channel dimension. This innovative method results in higher training throughputs and improved top-1 accuracy, offering a more sustainable baseline for future architecture designs.
AUTO-GPT
Auto-GPT is an experimental open-source application showcasing the capabilities of the GPT-4 language model. This program, driven by GPT-4, autonomously develops and manages businesses to increase net worth. As one of the first examples of GPT-4 running fully autonomously, Auto-GPT pushes the boundaries of what is possible with AI.
MARVIN
Marvin is a library for building AI-powered software.
AI IS ENTERING AN ERA OF CORPORATE CONTROL
PROMPT ENGINEERS ARE MAKING A LOT OF MONEY
AI PREDICTS ENZYME FUNCTIONS BETTER THAN LEADING TOOLS
GPT-4 IS MORE ACCURATE WHEN ASKED TO CRITIQUE ITSELF
Researchers have found that asking GPT-4 to critique its own work leads to a 30% performance boost.
HONG KONG’S CRYPTO AMBITIONS GET A BOOST FROM U.S. CRACKDOWN
Hong Kong has been ramping up its efforts to attract cryptocurrency companies as the US intensifies its regulatory crackdown. Prior to this, competition from Singapore and concerns about China's approach to crypto had caused Hong Kong to lose some ground as a crypto hub. The city is now trying to regain its status by offering a more favorable regulatory environment. Hong Kong's Securities and Futures Commission has proposed a new licensing framework focused on investor protection. This effort has gained some traction as 20 crypto companies have indicated plans to move to Hong Kong with 80 firms interested in doing so.
DUNESQL OFFICIALLY LAUNCHED
Dune has officially launched DuneSQL, aimed at creating the fastest and most convenient crypto data querying experience. DuneSQL, based on the open-source query engine Trino, will eventually replace Spark SQL and Postgres. DuneSQL provides blockchain-specific enhancements, and its roadmap includes materialized views, user-defined functions, querying personal data sources, and more. To help users migrate to DuneSQL, an automated tool has been created that will prompt users to move queries from Dune Engine V2.