Blogroll
When I'm not with my family or writing I'm probably browsing one of this blogs. If your time is limited I recommend you to skip my blog and read one of these ones instead. This is a list of all the posts published by my favourite blogs during the last 30 days.- 2026-08-03 - Beyond Bigger MoE: How Kimi K3 Scales Context, Depth, and Agents
- 2026-08-01 - Let’s Play “htmx 4: the game”
- 2026-08-01 - The Probe
- 2026-07-31 - Extracting “Ocarina of Time: Master Quest” ROM from the Legend of Zelda: Wind Waker bonus disc
- 2026-07-30 - Coming of a New Sun
- 2026-07-29 - From CUDA to MLX: How K-Search Brings Decades of Kernel Expertise to Apple Silicon
- 2026-07-29 - Cloudflare Deployment Modes for a CRUD App
- 2026-07-29 - Vectors Week: become a savier vector db customer
- 2026-07-29 - Just brute force your embeddings
- 2026-07-28 - LLMs Will Benefit from Scratch Workspaces
- 2026-07-28 - Kimi K3 Architecture Notes
- 2026-07-27 - Data leakage and independence
- 2026-07-26 - A Few Notable Open-Weight Models This Week
- 2026-07-26 - Teaching LLMs to Update Beliefs for Efficient Long-Horizon Interaction
- 2026-07-26 - What Is the Inverse Covariance Matrix?
- 2026-07-26 - How can LLM RL Work Despite Information-Theoretic Inefficiency
- 2026-07-25 - Correction for Listing 6.5 in Build a Reasoning Model From Scratch
- 2026-07-24 - Book Review: Python for Algorithmic Trading Cookbook
- 2026-07-24 - Principal Component Analysis: an embedding shrink-ray
- 2026-07-23 - Maybe Jacobian Conjecture Counterexamples are not Rare
- 2026-07-23 - LLMs break down in funny ways when told the Jacobian Conjecture counterargument
- 2026-07-23 - Exercises in benchmarking and evals, part 7: DeepSWE, Senior SWE-Bench, napkin math, and winter tires
- 2026-07-23 - claude gets a body #1
- 2026-07-22 - PAL GameCube haul from Kraków, Poland (EuroPython 2026)
- 2026-07-20 - Harness Handbook: The Missing Layer for Editing AI Agents
- 2026-07-20 - What's next for River (and myself)
- 2026-07-20 - Class of One: Podcast feed from NotebookLM summaries
- 2026-07-18 - What's the deal with all the random weekly quota resets for agents lately?
- 2026-07-18 - Controlling Reasoning Effort in LLMs
- 2026-07-18 - Point-in-Time Joins
- 2026-07-17 - Which Tech CEOs Are Gamers?
- 2026-07-17 - Book Review: Time Series with PyTorch
- 2026-07-17 - Is Endogenous Exploration Sufficient?
- 2026-07-16 - Inkling: A New Open-Weight 975B MoE with a Few Surprises
- 2026-07-16 - Habits
- 2026-07-16 - EuroPython 2026: Learning from the “not-so-secret” Python security cabal
- 2026-07-13 - Don't ask what you want. Ask who you want to be
- 2026-07-13 - Autarkic Agency Will Likely Migrate Upwards
- 2026-07-12 - 200,000 Subscribers
- 2026-07-12 - Baudrillard and Interiority
- 2026-07-11 - Generated and suppressed demand.
- 2026-07-11 - Do Automated Evals Work?
- 2026-07-11 - Make no assumptions.
- 2026-07-09 - GPT 5.6 Has 72 Possible Configurations. What's A Good Default?
- 2026-07-09 - Starting and iterating on a Kaggle competition in Google Antigravity
- 2026-07-09 - Jiggy: a desktop agent wearing a slime costume
- 2026-07-09 - Why write code in 2026
- 2026-07-08 - Now Go Build CTO Fellowship: Season 2
- 2026-07-08 - Cayley graph search with Claude Code: what puzzle competitions look like in 2026
- 2026-07-08 - Ephemeral Agent Workbenches
- 2026-07-07 - Intelligence is Free, Now What? Data Systems for, of, and by Agents
- 2026-07-07 - Playing the “Second Quest” of Legend of Zelda: Four Swords Adventures
- 2026-07-06 - Make invalid states unrepresentable (for your agents)
- 2026-07-06 - Book Review: GPU-Accelerated Computing with Python 3 and CUDA