Topics

17 shown
intermediate Reinforcement Learning

Agent Reward Hacking

How autonomous AI agents exploit poorly secured evaluation infrastructure, and what zero-trust testing pipelines look like.

@Marcus-Wee
intermediate Deep Learning

Autoencoders

From deterministic autoencoders to variational and vector-quantised models, and the gradient-estimation tricks (reparameterisation, Gumbel-Softmax, REINFORCE) that make them trainable.

@Zak-T
beginner Classical ML

Bayesian View on Regularization

Understanding ridge and lasso regularization through Bayesian priors and maximum a posteriori estimation.

@CJBuzz
beginner Classical ML

Classical Interpretability Techniques

A survey of classical and model-agnostic techniques for understanding how machine learning models make predictions.

@Praneeth-Suresh
beginner Classical ML

Gradient Descent

Iterative first-order optimisation algorithm for differentiable loss functions.

@Praneeth-Suresh, @N00bcak
intermediate World Modelling

Joint-Embedding Predictive Architecture

An introduction to how JEPA learns by predicting latent representations, with examples from I-JEPA and action-conditioned world models.

@Kia-Lok
beginner Classical ML

Linear Regression

Fitting a linear model to data using least squares optimisation.

@Praneeth-Suresh
intermediate Deep Learning

Neural Networks

Multi-layer parametric function approximators trained via backpropagation.

@Praneeth-Suresh
intermediate Generative

PEFT

Parameter-efficient fine-tuning — adapt large models on a single GPU by freezing the base weights and training a small delta.

@Marcus-Wee
beginner Classical ML

Probability And Bayesian Inference

the bayesian perspective of things

@checkpoint214159
intermediate Reinforcement Learning

Q-Learning

Model-free reinforcement learning algorithm that learns optimal action-value functions.

@Praneeth-Suresh
advanced World Modelling

Recurrent State Space Models (RSSM)

An introduction to recurrent state-space models, their training objective, and their evolution through PlaNet, Dreamer, MuDreamer, and R2-Dreamer.

@Kia-Lok
intermediate World Modelling

State Space Models for Sequence Modelling

An introduction to state space models, their continuous and discrete forms, recurrent and convolutional representations, HiPPO, S4, S5, and Mamba.

@Kia-Lok
intermediate Classical ML

Support Vector Machines

Maximum-margin classifiers using kernel methods for non-linear decision boundaries.

@Praneeth-Suresh
advanced Generative

Transformers

Self-attention-based sequence models that power modern LLMs and vision models.

@Praneeth-Suresh
intermediate Reinforcement Learning

Why you should care about PPO

why did ppo become the goat

@checkpoint214159
intermediate World Modelling

World Models

Learned environment models that let agents predict, plan, and learn from imagined futures.

@N00bcak