01May 6, 2026/Jamie Dborin

Tensor Network Attention

A note on how to interpret different attention mechanisms as tensor networks and how you can use that to inform inference and training time kernel decisions.

02May 5, 2026/Fergus Finn

In search of wasted bits: how much information do LLM weights carry?

An empirical investigation into the byte-level entropy of model weights across numeric formats and model families.

03April 27, 2026/Fergus Finn

tANS: precomputing rANS

An intro to tANS: a table-based entropy coder that removes rANS's per-symbol division while keeping Shannon-optimal compression.

04April 21, 2026/Fergus Finn

Also-rANS: Asymmetric Numeral Systems for entropy coding

An intro to rANS: an entropy coding method for losslessly encoding & decoding streams of bytes quickly.

05April 6, 2026/Fergus Finn

70x faster cold(ish) starts for SGLang

Checkpoint/restore with CRIU and cuda-checkpoint, from 12 minutes to 10 seconds on a B200.

06March 23, 2026/Jamie Dborin

OCR and the Bitter Lesson

Recently released open source OCR models are starting to replace expert-based OCR systems. This post walks through an evaluation exercise of specialist and general OCR agents.

07February 17, 2026/Fergus Finn

Weighted random fallback flattens to uniform under high error rates

When a weighted-random fallback rejects samples and retries without replacement, high error rates cause low-weight models to be selected far more often than their weights suggest.

08January 22, 2026/Jamie Dborin

QueueSpec: Drafting While You Wait

Speculative decoding speeds up LLM generation by letting a system propose several “draft” tokens at once, and then having the target model verify them in a single forward pass. The usual question is: where do we get good drafts cheaply? In this post, we explore queue speculation (QueueSpec): draft tokens come from a smaller model that runs while a request is queuing, so verification can start immediately once the request is serviced. At doubleword we use speculative decoding techniques like this and other throughput-specific optimizations to deliver cheaper inference at scale, by sacrificing end to end latency. If you want to get started with some free credits sign up here: Doubleword Platform

09January 22, 2026/Meryem Arik

Price Reduction for Qwen3-235B on Doubleword

Today we’re reducing the price of our highest-intelligence model, Qwen3-235B-A22B-Instruct.

10January 16, 2026/Fergus Finn

Scaling Curation with LLM Comparisons

Building a content discovery system using parallel primitives and BST-based ranking with LLM comparisons

11January 13, 2026/Fergus Finn

LLM powered data structures: A lock-free binary search tree

A lock-free binary search tree optimized for expensive async comparisons, with threaded linked list for O(1) sorted iteration

12January 8, 2026/Jamie Dborin

ZeroDP: Just-In-Time Weight Offloading over NVLink for Data Parallelism

High throughput inference of LLMs using JIT weight offloading to optimize KV Cache.