TLDR Index
TLDR AI

TLDR AI 2024-01-05

5 sections16 entrieslinks blockcomplete parse

Parse notes

  • Subscription/account footer truncated before parsefooter_truncated

01HEADLINES & LAUNCHES

PERPLEXITY RAISES $73.6M

The generative search company, which has seen massive adoption in tech circles, has raised another round of funding from top VCs.

reuters.com1 min read

OPENAI’S GPT STORE WILL LAUNCH NEXT WEEK

OpenAI plans to launch its GPT Store sometime in the coming week. To be eligible for listing, developers have to ensure their GPTs are compliant, verify their user profile, and make their creations “public”.

techcrunch.com2 min read

GOOGLE APPEARS TO BE WORKING ON A PAID ADVANCED VERSION OF BARD

Google is reportedly developing "Bard Advanced," an enhanced version of its Bard AI tool that will be accessible through a paid Google One subscription. This upgraded version, powered by Google's top-tier large language model Gemini Ultra, offers advanced math and reasoning capabilities compared to the current Bard running on Gemini Pro.

theverge.com2 min read

02RESEARCH & INNOVATION

OPEN MUSE REPRODUCTION

MUSE is a masked image generation model that uses Transformers. HuggingFace and Stability AI have worked to reproduce it in the open and have released training code, model checkpoints, and a paper. The models were not trained for as long as some of the best image diffusion models, but they show promise!

huggingface.co21 min read

SEMANTIC SEGMENTATION FOR AUTONOMOUS DRIVING

Researchers have developed CAINet, a novel approach for improving RGB-T semantic segmentation, crucial for autonomous driving. This system uniquely combines different types of data, focusing on their complementary aspects and global context.

arxiv.org18 min read

3D-AWARE VISUAL QUESTION ANSWERING

While Visual question answering (VQA) has seen advancements, most models only tackle 2D reasoning, neglecting the complexity of 3D visual scenes. This research presents 3D-aware VQA.

arxiv.org21 min read

03ENGINEERING & RESOURCES

RAGATOUILLE LIBRARY FOR RETRIEVAL PIPELINES

Retrieval Augmented Generation (RAG) is a way to include external knowledge into a language model’s generation. This library allows for training and researching state-of-the-art RAG systems.

github.comGitHub repository

REAL-TIME VIDEO TRACKING

ODTrack, a new video tracking system, improves the way we track objects in videos. Unlike older systems that struggled with connecting past and future frames, ODTrack uses a method called 'online token propagation' to better link consecutive video frames.

github.comGitHub repository

POISONING ATTACKS AGAINST RECOMMENDER SYSTEMS

This survey provides an in-depth look at the vulnerability of recommender systems to 'poisoning attacks', where harmful data is intentionally added to manipulate results.

github.comGitHub repository

04MISCELLANEOUS

INTEL'S GAUDI 2 CHIP SURPASSES A100

The new chip from Intel, available on a number of platforms and on-premise deployment, has excellent training throughput, flops, and decoding bandwidth.

databricks.com4 min read

THE AI TRUST CRISIS

Dropbox's new AI features have raised privacy concerns among users who fear their data might be used to train OpenAI's models, despite Dropbox's denial and a policy that requires customer consent for such use. This situation highlights a broader crisis of trust in AI and data privacy, emphasizing the need for clear communication and transparency from companies about their data usage practices.

simonwillison.net8 min read