Best newsletters and posts about relevant ML concepts tag
10/29/2021 10:14:34 AM
The conclusion of our self-supervised learning series
4/26/2022 11:24:44 AM
In this issue: we overview Centralized vs. Decentralized Distributed Training Architectures; we explain GPipe, an Architecture for Training Large Scale Neural Networks; we explore TorchElastic, a
6/28/2022 11:14:26 AM
+ what GNNs on Dynamic Graphs; and the exploration of DeepMind's Jraph, a GNN Library for JAX.
6/21/2022 11:24:37 AM
In this issue: we explain Graph Convolutional Neural Networks; we overview the original GCN Paper; we explore PyTorch Geometric, one of the most complete GNN frameworks available today. Enjoy the
7/26/2022 11:14:27 AM
In this issue: we discuss what to test in ML models; we explain how Meta uses A/B testing to improve Facebook's newsfeed algorithm; we explore Meta's Ax, a framework for A/B testing in PyTorch.
8/23/2022 11:34:27 AM
In this issue: we start the new series about text-to-image models; we discuss CLIP, a neural network that can learn image representations while being trained using natural language datasets; we explore
8/25/2022 12:14:28 PM
The new model uses text-to-image and image-to-image generation to produce astonishing artistic outputs.
9/6/2022 11:14:27 AM
+OpenAI's GLIDE; +the Hugging Face text-to-image catalog
9/22/2022 12:04:30 PM
Recently Amazon Research published three papers about BERT-based models
9/20/2022 11:14:30 AM
+Google's Parti; +MS COCO
9/15/2022 12:04:31 PM
It combines compression and system optimization techniques for building smaller and more efficient deep learning architectures
9/27/2022 11:14:26 AM
+the original VQGAN+CLIP paper; +VQGAN+CLIP implementations
10/13/2022 11:54:31 AM
The new model builds on the principles of text-to-image methods to produce visually astonishing videos
10/18/2022 11:14:28 AM
In this issue: we explain Meta AI's Make-A-Scene; we discuss Meta AI's Make-A-Scene Paper; we explore LAION, one of the most complete training datasets for text-to-image synthesis models. Enjoy
10/20/2022 12:04:32 PM
The new AI agent was able to discover new algorithms in a super challenging field of matrix multiplication
11/22/2022 12:14:28 PM
+Manifold; +Meta's Captum
11/22/2022 1:14:27 PM
Preventing toxic content, reducing bias and memorization have been some of the main challenges faced by the DALL-E 2 team
11/17/2022 12:44:27 PM
ReAct provides an architecture that triggers actions based on language reasoning paths
12/22/2022 12:14:28 PM
The model fine tuned GPT-3 to improve its ability to follow instructions.
12/20/2022 12:14:27 PM
Partial dependence plots, interpretable time series forecasting and Google's fairness indicators.
1/20/2023 5:54:54 AM
The model is one of the most impressive achievements in self-supervised learning research to this day.
1/24/2023 12:15:01 PM
Counterfactual explanations as an ML interpretability method, Google's StylEx and Microsoft's DiCE implementation
1/20/2023 3:44:54 AM
LIME, Meta AI research on interpretable neurons and the Alibi Explain framework.
1/26/2023 12:14:29 PM
The new generative AI model shows significant efficiency improvements over models like Stable Diffusion, Imagen and Parti.
1/20/2023 3:04:54 AM
The model is actively used in NVIDIA's Omniverse platform.
1/31/2023 12:14:29 PM
Interpretability methods optimized for deep neural networks, OpenAI's interpretability technique to discover multimodal neurons on CLIP and the Eli5 framework.
1/10/2023 12:14:27 PM
SHAP method, MIT taxonomy for ML interpretability and BAIR's iModels framework.
2/2/2023 12:14:27 PM
One of the techniques that enable the ChatGPT breakthrough comes from a 2017 research paper.
1/5/2023 12:14:27 PM
Edge 258: Inside OpenAI's Point-E: The New Foundation Model Able to Generate 3D Representations from Language
The new model combines GLIDE with image-to-3D generation models is a very clever and efficient architecture.
1/3/2023 12:14:30 PM
Local model-agnostic interpretability, IBMs ProfWeight research and the InterpretML framework.
12/29/2022 12:14:29 PM
An overview of the AI techniques behind OpenAI's new supermodel
3/2/2023 12:14:31 PM
Claude uses an interesting technique called Constitutional AI to enable safer content.
12/27/2022 12:14:27 PM
ALE method, OpenAI Microscope and IBM's AI 360 Explainability Toolkit.
12/27/2022 3:44:27 AM
ALE method, OpenAI Microscope and IBM's AI 360 Explainability Toolkit.
12/13/2022 12:14:27 PM
Global model-agnostic interpretability, student-teacher intrepetability methods and the Lucid library.
3/21/2023 11:14:28 AM
Vertical federated learning, Google's research about using federated learning to optimize mobile keyword predictions and the Flower framework.
3/23/2023 11:14:27 AM
The model is significatively smaller than GPT-3.5 but matches its performance on many important LLM benchmarks.
3/16/2023 2:55:13 PM
Sparrow uses a combination of large language models and reinforcement learning to enable a safer conversational experience.
3/28/2023 11:17:36 AM
Federated transfer learning, the TorchFL paper and the OpenFL framework.
3/30/2023 11:14:41 AM
The model powering services such as Bard and the conversational capabilities in Google Suite.
12/1/2022 12:54:30 PM
Large pretrained models are changing the mechanics of intelligent applications
4/4/2023 11:14:32 AM
Cross-silo federated learning(FL), Amazon's research on personalized FL and IBM's FL framework.
11/29/2022 12:24:32 PM
In this issue: we classify ML interpretability methods; we explore the building blocks of interpretability by Google Research; we explain TensorWatch, an open-source framework for debugging ML models.
4/6/2023 11:24:38 AM
Inside Alpaca: The Language Model from Stanford University that can Follow Instructions and Match GPT-3.5
The model is based on Meta AI's LLaMA and remains significatively smaller than GPT-3.5.
4/11/2023 11:14:49 AM
Cross device federated learning(FL), Google's work on FL with differential privacy and the FedLab framework
4/20/2023 11:14:28 AM
Dolly builds on the principles of InstructGPT on the GPT-J model.
4/19/2023 6:55:28 AM
LangChain is part of a generation of new frameworks that are integrating LLMs into mainstream software development lifecycles.
4/25/2023 11:14:37 AM
A summary of the topics discussed in the last 8 weeks.
4/19/2023 5:24:50 AM
Applying deferential privacy to federated learning(FL) scenarios, Meta AI's research and the best open source frameworks in this area.
4/27/2023 11:14:48 AM
Created by researchers from UC Berkeley, CMU, Stanford, and UC San Diego, Vicuna is part of the new wave of models that use Meta's LLaMA as its foundation.
11/10/2022 1:04:30 PM
The new algorithm combines reinforcement learning and Monte Carlo tree search to show unique levels of mathematical reasoning
5/2/2023 5:34:29 PM
TheSequence Thank you for reading TheSequence. As a token of our appreciation, we're offering you a limited-time offer of 20% off a paid subscription. Redeem special offer Here are the benefits you
5/2/2023 11:14:32 AM
A new series about new generation foundation model methods, Anthropic's Constitutional AI paper and LangChain.
11/8/2022 12:14:26 PM
+NVIDIA's textual inversion approach; +Outpainting interfaces
5/4/2023 11:14:32 AM
Edge 288: Inside DeepSpeed-Chat: Microsoft’s New Framework to Create ChatGPT-Like Models Based on Human Feedback
The new framework builds on the scalability capabilities of DeepSpeed to fine tune LLMs using RLHF.
11/3/2022 12:04:29 PM
The Google's new text-to-video super model can generate high-frame fidelity videos from textual inputs
5/18/2023 11:15:36 AM
The framework was created by the collaboration of Togeter, LAION, and Ontocord.
5/23/2023 11:15:20 AM
Instruction following LLs, OpenAI's InstructGPT and the Dust LLM framework.
5/16/2023 11:14:27 AM
1) Reinforcement Learning with Human Feedback(RLHF) 2) The RLHF paper, 3) The transformer reinforcement learning framework.
5/25/2023 11:15:15 AM
Edge 294: Inside StarCoder: Hugging Face's New LLM that Can Generate Code in Over 80 Programming Languages
StarCoder was created by Hugging Face and ServiceNow as part of the BigCode project.
5/12/2023 7:16:04 AM
The model provides a lighter, open-source alternative to ChatGPT and includes EasyLM, a framework for training and fine-tuning LLMs.
5/9/2023 11:14:27 AM
Chain of thought prompting(CoTP), Google's original (CoTP) paper and the OpenChatKit framework
11/1/2022 11:14:27 AM
+retrieval augmented diffusion models; +Stable Diffusion interfaces
5/30/2023 11:14:31 AM
What if LLMs could auto improve their own instruction following capabilities?
6/1/2023 11:14:36 AM
The technique is one of the first attempts to utilize LLMs as a explainability foundation.
10/27/2022 11:55:08 AM
On Thursdays, we dive deep into one of the freshest research papers or technology frameworks that is worth your attention. Our goal is to keep you up to date with new developments in AI to complement
10/11/2022 11:14:28 AM
In this issue: we explain DALL-E 2; we discuss the DALL-E 2 paper; we explore DALL-E Mini (Now Craiyon), the most popular DALL-E implementation on the market. Enjoy the learning! 💡 ML Concept of the
6/20/2023 7:25:10 AM
The eCommerce giant published some details about the platform powering its ML workflows
6/20/2023 11:14:30 AM
The ideas for decoupling model knowledge from language generation.
6/20/2023 6:44:49 AM
The model quickly top the Open LLM Leaderboard that ranks the performance of open source LLMs.
6/22/2023 11:15:08 AM
The new suite of models was released by MosaicML and support models optimized for Instructions, Chats, Stories and More.
6/13/2023 11:14:28 AM
What are the different ways to augment LLMs with tools.
6/12/2023 3:44:55 AM
Can LLMs master knowledge tools?
6/27/2023 11:14:27 AM
What are the main types of techniques to augment LLMs with external information.
6/12/2023 2:44:43 AM
The model expands Vicuna with vision capabilities similar to BLIP-2 in one of the most interesting open source releases in the multi-modality space.
6/29/2023 11:26:39 AM
Edge 304: Inside AlphaDev: DeepMind’s Newest Breakthrough Model that Was Able to Discover New Computer Science Alg…
Built on the foundation created by AlphaZero, the model discovered new and improved existing sorting algorithms.