Latest

Every published item, in order · 4310 items
Oct 26
Oct 25
Tomorrow 08:00, the daily lands in your inboxTelegram channel or email. Same screening as this page, no ads.
How-toBuildersSelected20:01
Sebastian Raschka completes reasoning-from-scratch video series

Building a reasoning model directly from code demystifies how chain-of-thought-style systems actually work, knowledge that transferable to fine-tuning or evaluating production models. Practitioners and ML engineers who learn best by implementation can follow the series as a structured course. It comes from the author of widely used LLM from-scratch material, so the codebase will likely be maintained and referenced going forward.

Official accountSebastian Raschka@rasbtSource
How-toBuildersSelected02:19
Y Combinator Paper Club breaks down the full AI inference stack

The claim that serving consumes six times the training FLOPs reframes inference as the real cost centre, not model training. Builders running inference workloads can use the full-stack walkthrough to understand why the same open model behaves differently across providers and where compiler and caching gains come from. The segment on SGLang memory management is especially relevant for anyone serving agents with long contexts.

Y Combinator@ycombinatorSource
Oct 24
How-toBuildersSelected19:43
Billion-context: layered compression lets agents restore billion-token history

Unlike one-shot summaries, restoring original content on demand means day-one decisions remain traceable after weeks-long sessions with multiple refactors — no detail is permanently lost to compression. Developers running long-horizon coding or tool-using agents can adopt it across Claude Code, Codex, OpenCode and DeepSeek Harness to cut token costs while keeping history searchable.

AIGCLINK@aigclinkSource
Oct 23

Only the latest 2000 items are listed; for older ones see the daily archive