RESEARCH

Vision Transformers: Tokenizing Images Like Language Models

By

–

03 July 2023 17h40

In a tweet: similar to LLMs (language transformers) but tokenizing images instead of text:

→ View original post on X — @miketamir

MIT MAS.S68 Course on AI and Machine Learning

By

@miketamir

–

03 July 2023 17h37

MIT MAS.S68! https://
bit.ly/441rwBF #AI #MachineLearning #DeepLearning #LLMs #DataScience

Modern Classification: Multiple Output Nodes vs Sigmoid Threshold

By

–

03 July 2023 17h21

So, in this case, rather than using a threshold on a logistic sigmoid score [0, 1], in most modern architectures, you usually define at least 2 output nodes. Then, you determine the class based on which output node has the highest score.

Memory-Saving Techniques for Tabular Data Transformers

By

–

03 July 2023 17h14

but yeah, I am training classifiers in both cases, the same memory-saving techniques could potentially be applied to the recent transformers for tabular data.

→ View original post on X — @nathanbenaich

CERN Revolutionizes Spinout Strategy with Royalty-Based IP Licensing

By

@nathanbenaich

–

03 July 2023 17h07

Here’s @CERN
, Europe’s iconic and world-respected centre for nuclear research, rewriting its spinout playbook. 0% equity express agreement with 2% royalties if you monetise IP you licensed. KPIs set up for success. A star for @spinoutfyi
! Let’s go @onceageek
!

→ View original post on X — @marktabnet

Altman Addresses Safety Concerns Beyond GPT-4 Development

By

@marktabnet

–

03 July 2023 16h07

In April 2023, OpenAI CEO Sam Altman spoke at MIT's The Future of Business with AI event. He said, "We are doing other things on top of GPT-4 that I think have all sorts of safety issues that are important to address" Read More: https://
bit.ly/435FsK5

→ View original post on X — @learnopencv

SAM: Revolutionary Image Segmentation Model with Unmatched Precision

By

@learnopencv

–

03 July 2023 16h01

🎥New Video:https://t.co/MfidGJzK7w

Dive into the specifics of SAM in our latest video – a game-changing image segmentation model with unmatched precision. #AI #ImageSegmentation #AIInnovation #computervision pic.twitter.com/PBvhLh5BjL
— Satya Mallick (@LearnOpenCV) 3 juillet 2023

New Video: https://
youtube.com/watch?v=mtHEBU
dYRYU
… Dive into the specifics of SAM in our latest video – a game-changing image segmentation model with unmatched precision. #AI #ImageSegmentation #AIInnovation #computervision

ChatGPT’s Knowledge Cutoff Limitation and Synthetic Data Challenges

By

–

03 July 2023 15h57

3/3 Intuitively one major criticism of ChatGPT is it's 2020 knowledge cut-off. If you are only using synthetic data from old models, it'd be challenging to get more contemporary knowledge into the model.

Synthetic Data Generation Limitations in Predictive Model Improvement

By

–

03 July 2023 15h55

2/2 I think there's been a lot of research on generating synthetic data to boost predictive models (incl. classifiers). The concern is that you'll never be able to generate anything "new" by recycling models (trained on a fixed dataset) that generate synthetic data.