This two-legged robot taught itself how to walk blogs.mathworks.com/headlines/2021…

An Overview of Diffusion Models: Applications, Guided Generation, Statistical Rates and Optimization ift.tt/FJiwevm

Very excited to attend The AAMAS Conference this week! The Citizen-Centric Artificial Intelligence Systems team is presenting 2 full papers, 1 Blue Sky Ideas paper, 2 extended abstracts and 1 demo. More details below 👇

Presenting Tesfay Gebrekidan's (Tesfay Zemuy Gebrekidan) work on Deep Reinforcement Learning with Coalition Action Selection in Crystal Room 2 at The AAMAS Conference now (10:30-12:30 session). Hope to see you there!

My PhD thesis titled "Deep Reinforcement Learning for Online Combinatorial Resource Allocation with Arbitrary State and Action Spaces" is now available publicly. See the highlight on LinkedIn. linkedin.com/posts/tesfayz_…

[VAE] by Hand ✍️ A Variational Auto Encoder (VAE) learns the structure (mean and variance) of hidden features and generates new data from the learned structure. In contrast, GANs only learn to generate new data to fool a discriminator; they may not necessarily know the

Direct Preference Optimization (DPO) has become one of the go-to methods to align large language models (LLMs) more closely with user preferences. If you want to learn how it works, I coded it from scratch: github.com/rasbt/LLMs-fro…

Understanding full-RL requires understanding the difference between multiarmed bandit -- which has a single observation and multiple actions, contextual bandit (semi-RL) -- which has multiple observations, and the full-RL with delayed reward. #RLHF is not fully #RL.

AI is going to be the rational reviewer. At least it replaces the role of reviewers with minimal involvement of chairs. I am astonished with how cool ai and ml papers has summarized our work as linked here. aimodels.fyi/papers/arxiv/c… #AI #ML

arxiv.org/abs/2412.05265

A Deep Dive into Reasoning LLMs This is a really nice summary of the progress made in post-training and reasoning LLMs. Highly recommend this one!

growkudos.com/publications/1…

Open research: QBound: Environment-aware Q-value Bounds for Stable Temporal Difference Learning. linkedin.com/posts/tesfayz_…