Раздел базы данных · 12 записи
Исследования
The papers the modern stack is built on — from the transformer to scaling laws, RLHF and reasoning models.
Заметки с места событий
Дюжина PDF-файлов лежит в основе инфраструктуры стоимостью триллион долларов. Это статьи, где идеи впервые появились, — большинство удивительно доступны для чтения, и каждая запись здесь объясняет, что было важно, без формул.
Начните с этих трёх
AI
Attention Is All You Need (2017)
Восемь авторов, одна архитектура, всё остальное — потом.
TC
Training Compute-Optimal LLMs — Chinchilla (2022)
Поправка, изменившая то, как обучается каждая модель.
DR
DeepSeek-R1 — Reasoning via RL (2025)
Открытый рецепт рассуждений — опубликован, пока конкуренты держали свои в тайне.
Foundational papers
12
AI
Attention Is All You Need (2017)
The eight Google authors who invented the transformer.
A(
AlexNet (2012)
The ImageNet result that started the deep-learning era — on two gaming GPUs.
CA
Constitutional AI (2022)
Anthropic's method: models critique themselves against written principles.
DR
DeepSeek-R1 — Reasoning via RL (2025)
The open recipe for reasoning models — a Nature cover, from a hedge fund's lab.
DD
Denoising Diffusion Probabilistic Models (2020)
The DDPM paper behind every image and video generator.
F(
FlashAttention (2022)
The kernel trick that made long context affordable.
HA
Highly Accurate Protein Structure Prediction (2021)
AlphaFold 2 in Nature — the result that won a Nobel.
I—
InstructGPT — RLHF at Scale (2022)
How raw language models became assistants people can use.
LM
Language Models are Few-Shot Learners (2020)
The GPT-3 paper: capability from scale, tasks from prompts.
LL
LoRA: Low-Rank Adaptation (2021)
Fine-tuning for the GPU-poor — tiny adapters instead of full weights.
SL
Scaling Laws for Neural Language Models (2020)
Kaplan et al.: loss falls as a power law in compute — the investment thesis, quantified.
TC
Training Compute-Optimal LLMs — Chinchilla (2022)
DeepMind's correction: most models were undertrained on data.