@TensorTonic: If you want to actually understand LLMs (not just use them), read these in order: 1. Attention Is All You Need (transfo…
Summary
A Twitter thread recommending ten foundational papers and works to understand LLMs, from the original Transformer to DPO.
View Cached Full Text
Cached at: 07/02/26, 12:21 PM
If you want to actually understand LLMs (not just use them), read these in order:
- Attention Is All You Need (transformers)
- GPT-2 (scaling + zero-shot)
- Scaling Laws (Kaplan, 2020)
- GPT-3 (few-shot)
- Chinchilla (how much data you actually need)
- InstructGPT (RLHF, why ChatGPT works)
- LoRA (fine-tuning without going broke)
- FlashAttention (why it’s fast)
- Chain-of-Thought (reasoning)
- DPO (RLHF without the pain)
Similar Articles
How LLMs Actually Work (26 minute read)
A detailed walkthrough of how transformer-based LLMs work, covering tokenization, embeddings, attention, and next-token prediction without heavy math.
@techNmak: This is the best way to learn how LLMs work. Interactive. 3D. Step-by-step. Covers: → Embedding → Layer Norm → Self-Att…
An interactive 3D step-by-step guide to learning how LLMs work, covering key transformer concepts like embedding, self-attention, and softmax. It recommends a visual approach over reading papers.
@amitiitbhu: Research papers every LLM engineer must read: - Attention Is All You Need - BERT - GPT-3: Language Models are Few-Shot …
A list of essential research papers for LLM engineers, including key works on transformers, scaling laws, and fine-tuning techniques.
@TheAhmadOsman: INCREDIBLE RESOURCE The MOST COMPLETE GUIDE for understanding LLMs from first principles is now available online to rea…
A comprehensive free guide explaining LLMs from first principles, covering tokens, transformers, attention, fine-tuning, and local deployment.
@Alacritic_Super: If you want to master LLM inference, start with these three papers. They introduced many of the ideas powering today's …
This thread recommends three key papers for mastering LLM inference: PagedAttention, Sarathi-Serve, and SGLang, which introduce efficient memory management, chunked prefills, and structured generation techniques used in modern inference engines like vLLM and TensorRT-LLM.