# New? Start Here

The easiest way to find something that you’ll be interested in is to look at the archives and browse the titles, which (I hope) are descriptive. Using the built-in Google site search there would also be useful.

If you’re interested in knowing more about **graduate-level classes at
Berkeley**, I write reviews on all the ones I have taken. Here they are:

- CS 182/282A, Deep Neural Networks (GSI/TA, first time, second time)
- CS 267, Applications of Parallel Computing
- CS 280, Computer Vision
- CS 281A, Statistical Learning Theory
- CS 287, Advanced Robotics
- CS 288, Natural Language Processing
- CS 294-112, Deep Reinforcement Learning
- CS 294-112, Deep Reinforcement Learning (self-study)
- CS 294-115, Algorithmic Human-Robot Interaction
- CS 294-131, Special Topics in Deep Learning
- EE 227BT, Convex Optimization
- EE 227C, Convex Optimization and Approximation
- STAT 210A, Theoretical Statistics (Classical)
- STAT 210B, Theoretical Statistics (Modern)

I also write a lot technical subjects. For example, here are some **Generic
Technical Guides**, not including those related to my prelims studying
(discussed later):

- Ten Things Python Programmers Should Know (my most popular blog post)
- Better Saving and Logging for Research Experiments
- Adversarial Examples and Taxonomies
- Basics of Bayesian Neural Networks
- Independent Component Analysis — A Gentle Introduction
- Mathematical Tricks Commonly Used in Machine Learning and Statistics
- Going Deeper Into Reinforcement Learning: Fundamentals of Policy Gradients
- Going Deeper Into Reinforcement Learning: Deep-Q-Networks
- Going Deeper Into Reinforcement Learning: Q-Learning and Linear Function Approximation
- Understanding Backpropagation Gradient Computation: post 1 and post 2
- Domain Randomization Tips
- Importance Sampling Review
- Understanding Generative Adversarial Networks
- Fundamentals of Robotics (SE, Twists, etc.): post 1 and post 2

Here are notes on **Specific Research Papers** (for others, see this GitHub
repository), arranged roughly in order of topic similarity.

- (Asynchronous) Advantage Actor-Critic
- Prioritized Experience Replay
- Notes on the Generalized Advantage Estimation Paper
- Deep Q-learning from Demonstrations (DQfD)
- DQfD Follow-ups (3 papers: Distributed PER, Ape-X, Kickstarting)
- Off-Policy Deep Reinforcement Learning without Exploration
- Learning to Teach
- Born Again Neural Networks
- Papers That Have Cited Policy Distillation (links to external PDF)
- Interpretable and Pedagogical Examples
- Model-Agnostic Meta Learning (MAML)
- One-Shot Visual Imitation Learning via Meta-Learning
- Zero-Shot Visual Imitation
- Hindsight Experience Replay
- Overcoming Exploration in Reinforcement Learning with Demonstrations
- Learning to Act by Predicting the Future
- Learning to Poke by Poking: Experiental Learning of Intuitive Physics
- Self-Supervision Papers (7 papers, including some discussed separately)
- Dense Object Descriptors with rope and fitting applications (4-5 papers)
- Deep Learning for Robotic Grasping (3 papers)
- Read-Through of Multi-Level Discovery of Deep Options
- Understanding Deep Learning Requires Re-Thinking Generalization
- Transformer Networks for State of the Art Natural Language Processing
- Some Recent Results on Minibatch Markov Chain Monte Carlo Methods
- Understanding and Categorizing Scalable MCMC and MH Papers (7 papers)
- Minibatch Metropolis-Hastings (one of my papers!)

When I was preparing for **the AI prelims at Berkeley** (required for PhD
students), I wrote a lot about AI topics. I also wrote a “transcript” of my
prelims.

- My Prelims
*[Transcript]* - Miscellaneous Prelim Review (Part 1)
- Miscellaneous Prelim Review (Part 2)
- Markov Decision Processes and Reinforcement Learning
- Perceptrons, SVMs, and Kernel Methods
- Notes on Exact Inference in Graphical Models
- Closing Thoughts on Graphical Models
- The Least Mean Squares Algorithm
- Expectation-Maximization
- Hidden Markov Models and Particle Filtering
- Reading Russell and Norvig
- Stanford’s Linear Algebra Review

I later wrote a transcript of my qualifying exam.

I also blog about **academic conferences**. Here are the ones I have discussed,
in chronological order and the city:

- UAI 2017. Sydney, Australia
- ICRA 2018. Brisbane, Australia
- IJCAI 2018. Stockholm, Sweden
- ICRA 2019. Montreal, Canada
- ISRR 2019. Hanoi, Vietnam
- NeurIPS 2019. Vancouver, Canada

I sometimes make one post per day. For simplicity, the above will link to the
*first* blog post for each of the conferences.

If you are interested in knowing about **what it’s like being deaf** then, while
I obviously can’t claim to speak for everyone who has hearing impairments, here
are a few that might be informative:

- The Obligatory “Can I Lip Read?” Question
- The BVLC (BAIR) Retreat: Disaster Averted!
- Advocate for Yourself
- After a Few Weeks of CART, Why do I Feel Dissatisfied?
- The Problem with Seminars
- My Pre-College Education as a Deaf Mainstreamed Student
- New Closed-Captioning Glasses
- Hearing Aids: How They Help and How They Fall Short in Group Situations
- Technical Term Dilemma
- Why Computer Science is a Good Major for Deaf Students
- My Thoughts on CS 231n Being Forced To Take Down Videos

In addition, most of my posts on going to academic conferences discuss some of the above topics, as well as my prelims and quals transcripts.

Finally, I write sometimes about **the books I read**, such as in the following:

- All the Books I Read in 2019, Plus My Thoughts
- All the Books I Read in 2018, Plus My Thoughts
- All the Books I Read in 2017, Plus My Thoughts
- All the Books I Read in 2016, Plus My Thoughts
- Thoughts on How to Win Friends and Influence People
- Alan Turing: The Enigma
- The Master Algorithm: How the Quest for the Ultimate Learning Machine Will Remake Our World
- My Three Favorite Books I Read in 2015
- Dual Book Discussion on Political Development and Faith