Sezione del database · 12 record
Ricerca
The papers the modern stack is built on — from the transformer to scaling laws, RLHF and reasoning models.
Note sul campo
Una dozzina di PDF reggono un trilione di dollari di infrastrutture. Sono i paper in cui le idee sono comparse per la prima volta — la maggior parte è sorprendentemente accessibile, e ogni scheda qui spiega cosa ha contato, senza la matematica.
Inizia da questi tre
AI
Attention Is All You Need (2017)
Otto autori, un'architettura, tutto il resto da allora.
TC
Training Compute-Optimal LLMs — Chinchilla (2022)
La correzione che ha cambiato il modo in cui ogni modello viene addestrato.
DR
DeepSeek-R1 — Reasoning via RL (2025)
La ricetta aperta per il ragionamento — pubblicata mentre i rivali tenevano la loro segreta.
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.