AI, without the fog
Dispatches from inside the machine
Independent notes on AI research, local models, agent design, and the security decisions behind them. Pick a subject or search the archive.
Search articles →Browse topics
All articles →Latest articles
All articles →-
2026-10-19
Citations Are Not Evidence: Auditing What a RAG Answer Actually Claims
A bracketed [1] tells you where to look, not what you will find. How to check, claim by claim, whether a cited passage supports a generated answer.
AI research -
2026-10-17
Tokens Per Second Is Two Benchmarks Wearing a Trench Coat
A single tokens-per-second score hides prompt ingestion, decoding, and the workload that actually matters. Benchmark both halves or stop comparing machines.
Local models -
2026-10-16
Structured Output Is Not Truth: What a JSON Schema Can—and Cannot—Guarantee
A grammar can constrain a model's output shape. It cannot establish whether the answer matches the source. Here is a small, testable example.
Local models -
2026-10-14
Diffusion Language Models Have a Cache Problem. That Is Finally Getting Interesting
Diffusion language models can generate text in parallel, but repeated denoising makes them slow. New cache-aware work makes the trade-off worth watching.
Local models -
2026-10-13
Decision Models Are Eating Chat Models' Lunch
Cloudflare, Amazon, and Jev all shipped open decision models in October 2026. The question isn't which is best — it's why you'd use a chat model for routing in 2026.
Local models -
2026-10-12
Diffusion Language Models Have a Cache Problem. That Is Finally Getting Interesting
Diffusion language models can generate text in parallel, but repeated denoising makes them slow. New cache-aware work makes the trade-off worth watching.
Local models -
2026-10-09
Prompt Injection Is a Permission Problem
Filters help, but the durable defence against prompt injection is to make a manipulated agent unable to cause expensive harm.
AI security -
2026-10-07
NVIDIA's Tri-Mode Model Doesn't Just Draft — It Switches Gears
NVIDIA's Nemotron-Labs-Diffusion unifies AR, diffusion, and self-speculation in one architecture. It can switch modes mid-generation. The paper claims 6× more tokens per forward than AR — but the real story is how tri-mode changes what hardware you need.
AI research