MohanadOdema / Gen-AI-reads

Reads and links of interests for advancements in generative AI and LLMs

Gen-AI-reads

Reads and links of interests for topics of interest in the generative AI field

Contents

• Training
• Reasoning
• Compression
• Trustworthiness

Training

Scaling Laws for Large Language Models (LLMs)

• Chinchilla Scaling Laws for Large Language Models (LLMs) - High-level overview article by medium
• Training Compute-Optimal Large Language Models - DeepMind's Chinchilla Scaling Laws, 2022
• Scaling Laws for Neural Language Models - OpenAI's OG scaling law, 2020

Reasoning

Chain of Thoughts

• Chain-of-Thought Prompting Elicits Reasoning in Large Language Models - OG by Google Research, Brain Team, Jan 2023
• Tree of Thoughts: Deliberate Problem Solving with Large Language Models - Generalizing Chain of Thoughts, Deepmind @ Neurips 2023
• Graph of Thoughts: Solving Elaborate Problems with Large Language Models - Graphing the tree, 2023

World Models

• Recurrent World Models Facilitate Policy Evolution - World models explanation and demos, Neurips 2018
• Scaling GAIA-1: 9-billion parameter generative world model for autonomous driving - Wayve's generative world model for autonomous driving, 2023

Compression

Generative Image Compression

• High-Fidelity Generative Image Compression - Generative compression methods for high quality reconstructions, Neurips 2020

Trustworthiness

• DECODINGTRUST: A Comprehensive Assessment of Trustworthiness in GPT Models - Neurips'23 Best Paper Award on toxicity, stereotype bias, adv robustness, OOD, robustness against adv demonstrations, privacy, machine ethics, fairness, 2023