Foundations
Understand language models and choose a learning path.
- Llm Fundamentals (139)
- Courses And Learning Paths (179)
- From Scratch Implementations (1)
- Practitioner Overviews (387)
706 resources · page 22 of 23
Relational inductive biases, deep learning, and graph networks
Creator not supplied; see original source.
arXiv paper (2018) "Relational inductive biases, deep learning, and graph networks", defining the graph network block that unifies many GNN variants.
ResNet
Creator not supplied; see original source.
arXiv paper (2015) "Deep Residual Learning for Image Recognition", introducing ResNet residual networks.
RL4CO
Creator not supplied; see original source.
RL4CO documentation: reinforcement-learning benchmark and library for combinatorial optimisation problems.
RWKV
Creator not supplied; see original source.
arXiv paper (2023) introducing RWKV, a recurrent architecture trained in parallel like a transformer.
Efficiently Modeling Long Sequences with Structured State Spaces (S4)
Creator not supplied; see original source.
arXiv paper (2021) "Efficiently Modeling Long Sequences with Structured State Spaces", introducing S4, structured state space models for long sequences.
SAC
Creator not supplied; see original source.
arXiv paper (2018) introducing Soft Actor-Critic (SAC), off-policy maximum-entropy deep reinforcement learning.
SAM
Creator not supplied; see original source.
arXiv paper (2023) introducing Segment Anything (SAM), a promptable image segmentation model and dataset.
Scaling Laws for Neural Language Models
Creator not supplied; see original source.
arXiv paper (2020) "Scaling Laws for Neural Language Models", measuring power-law scaling of language model loss with size, data and compute.
scvi-tools
Creator not supplied; see original source.
scvi-tools: library of probabilistic deep generative models for single-cell omics data.
Switch Transformers
Creator not supplied; see original source.
arXiv paper (2021) introducing Switch Transformers, sparse mixture-of-experts models scaled to a trillion parameters.
A decoder-only foundation model for time-series forecasting (TimesFM)
Creator not supplied; see original source.
arXiv paper (2023) introducing TimesFM, a decoder-only foundation model for time-series forecasting.
tslearn
Creator not supplied; see original source.
tslearn documentation: Python machine learning tools for time-series analysis.
U-Net
Creator not supplied; see original source.
arXiv paper (2015) "U-Net: Convolutional Networks for Biomedical Image Segmentation", introducing the encoder-decoder architecture with skip connections.
Umar Jamil (YouTube)
Creator not supplied; see original source.
Umar Jamil's YouTube channel: line-by-line implementations of Transformer, LLaMA and diffusion models.
verifiers v1
Creator not supplied; see original source.
Prime Intellect post on verifiers v1, splitting environments into tasksets, harnesses and runtimes for agentic RL and evals.
VGG
Creator not supplied; see original source.
arXiv PDF of the VGG paper (2014) on very deep convolutional networks for large-scale image recognition.
ViT
Creator not supplied; see original source.
arXiv paper (2020) introducing the Vision Transformer (ViT) for image recognition.
WaveNet
Creator not supplied; see original source.
arXiv paper (2016) "WaveNet: A Generative Model for Raw Audio", introducing WaveNet, an autoregressive generative model of raw audio.
Wide & Deep
Creator not supplied; see original source.
arXiv paper (2016) "Wide & Deep Learning for Recommender Systems", introducing Google's Wide & Deep recommender model.
xLSTM (NXAI, Linz)
Creator not supplied; see original source.
arXiv paper (2024) introducing xLSTM, extended long short-term memory with exponential gating and matrix memory.
YOLO
Creator not supplied; see original source.
arXiv paper (2015) "You Only Look Once: Unified, Real-Time Object Detection", introducing YOLO, a single-stage real-time object detector.
AWQ
Creator not supplied; see original source.
arXiv paper (2023) introducing AWQ, activation-aware weight quantization for LLM compression.
EASA AI Concept Paper Proposed Issue 03
Creator not supplied; see original source.
EASA AI Concept Paper Proposed Issue 03 (PDF), extending aviation AI assurance guidance to more methods and Level 3 automation.
ELIS 2025 report (European Language Industry Survey)
Creator not supplied; see original source.
European Language Industry Survey 2025 report (PDF) on trends, including AI, in the translation industry.
GPTQ
Creator not supplied; see original source.
arXiv paper (2022) introducing GPTQ, accurate post-training weight quantization for generative transformers.
HBS: Navigating the Jagged Technological Frontier (BCG study)
Creator not supplied; see original source.
Harvard Business School working paper Navigating the Jagged Technological Frontier: field experiment with BCG consultants using GPT-4.
Reward Hacking in Reinforcement Learning (Lil'Log)
Creator not supplied; see original source.
Lilian Weng's survey of reward hacking in reinforcement learning, including in RLHF and coding tasks.
Machine Learning Street Talk (YouTube)
Creator not supplied; see original source.
Machine Learning Street Talk YouTube channel: long interviews and debates with AI researchers.
Mamba
Creator not supplied; see original source.
arXiv paper (2023) introducing Mamba, linear-time sequence modelling with selective state spaces.
Mem0 paper
Creator not supplied; see original source.
arXiv paper (2025) introducing Mem0, scalable long-term memory for production AI agents, with benchmarks.
Olmo 3 technical report
Creator not supplied; see original source.
arXiv paper (2025) Olmo 3 technical report on Ai2's fully open 7B and 32B models.
Smol Training Playbook (Hugging Face Space)
Creator not supplied; see original source.
Hugging Face Smol Training Playbook: end-to-end account of training the SmolLM3 model, including failures and decisions.