A QUIET PDF READER
01. AI
- Hawkeye Hardware-Aware GPU Kernel Optimization with Minimal Supervision
47P - Computer Use at the Edge of the Statistical Precipice
24P - ISCA 2023 Camera Ready TPUv4 Paper
14P - HowToScaleYourModel
141P - composer2
23P - Ribbit Identity Letter (2024, alternate 22-page version)
22P - Ribbit Pulse Q1 (2026)
47P - Ribbit Pulse Q4 (2025)
39P - Ribbit Pulse - Opus 4.5 Moment (February 2026)
29P - Ribbit Pulse Q2 (2025)
35P - Ribbit Pulse Q4 (2024)
34P - Ribbit Pulse Q3 (2024)
33P - Ribbit B2B Finance Letter (2021)
31P - Ribbit Identity Letter (2024)
23P - Ribbit China Letter (2019)
26P - Ribbit Wealth Management Letter (2017)
34P - Ribbit Crypto Letter (2018)
45P - Ribbit Insurance Letter (2014)
26P - Ribbit Bitcoin Letter (2013)
19P - Ribbit Token Letter (2025)
41P - Axe: A Simple Unified Layout Abstraction for Machine Learning Compilers
24P - Linear Layouts: Robust Code Generation of Efficient Tensor Computation Using $\mathbb{F}_2$
18P - Mind the Gap: Attainable Data Movement and Operational Intensity Bounds for Tensor Algorithms
17P - The Metropolis–Hastings algorithm
15P - Aristotle: IMO-level Automated Theorem Proving
23P - ECE 20869 HW0 - Getting to Know You
1P - PAT: a new algorithm for all-gather and reduce-scatter operations at scale
11P - Mathematics in the age of AI
13P - baseten inference stack
37P - spec dec
13P - ribbit power letter
42P - laguna-m1-xs2
37P - Kevin: Multi-Turn RL for Generating CUDA Kernels
27P - Foundations of Large Language Models
277P - Reinforcement Learning from Human Feedback
239P - The Scaling Era: An Oral History of AI, 2019–2025
400P - k3 tech report
47P - Efficiently Scaling Transformer Inference
19P - byol
35P - deepseek v4
58P - Efficient Memory Management for Large Language Model Serving with PagedAttention
16P - Reinforcement Learning from Human Feedback
225P - flash attn 3
22P - flash attn 2
14P - flash attn
34P - Kimi K2.5: Visual Agentic Intelligence
30P - DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
86P - deepseek v3
53P - Compiling machine learning programs via high-level tracing
3P - 1 bit llm
8P - reiner
37P - jeff dean thesis
34P - attention-is-all-you-need
15P