What is mattwood.fyi?

A live, structured feed of riffs and links from Matt Wood, Chief AI & Technology Officer at AWS — readable by people and queryable by agents as a typed knowledge graph.

Is there an API and is it free?

Yes. A public, read-only query API is available with no authentication and no API keys. See https://mattwood.fyi/openapi.json and https://mattwood.fyi/docs.

How should an AI agent use this site?

Fetch https://mattwood.fyi/feed.json for the full structured feed, or query the API for keyword/semantic search and the connection graph. See https://mattwood.fyi/agents/ for when-to-use guidance.

Clef: Open-Source Decision Models and RL Fine-Tuning

Connections

Related to

Kev: Tiny Decision Models on Qwen Both Clef and Kev focus on small/decision models for AI applications, with Kev specifically targeting tiny decision models on Qwen and Clef providing an open-source decision model platform — directly parallel approaches
Introducing System One Models and Jev System One Models and Jev also introduces specialized decision/inference models, making both items part of the same emerging category of purpose-built decision model platforms
Engineering a Calibrated Classifier Model Both address the engineering of calibrated decision-making in AI systems — Clef through RL fine-tuning of decision models, and the calibrated classifier paper through probabilistic output calibration

Supports

AX: Agentic Task Execution Platform Clef's open-source decision models and RL fine-tuning capabilities provide foundational infrastructure that agentic task execution platforms like AX can leverage for policy-driven agent decision-making
Automating Eval Design and Hillclimbing with Claude Clef's RL fine-tuning capabilities directly enable the kind of automated evaluation and hill-climbing workflows described in the Claude-based eval design automation, as RL fine-tuning depends on robust evaluation loops
Thinking Before Thinking: Scaling Agentic Inference Through Meta-Reasoning Clef's decision models and RL fine-tuning infrastructure supports the meta-reasoning and agentic inference scaling described in Thinking Before Thinking, as RL-trained decision models are key to efficient inference-time reasoning allocation

Type:	link
First published:	Thursday, 1 October 2026 at 18:02 UTC
Location:	https://blog.cloudflare.com/clef-decision-models/
URL:	https://mattwood.fyi/i/9654f595

I'm Matt Wood, and this is For Your Information. A live list of riffs and links for you and your agent, drawn from what I'm reading, noticing, questioning, concluding, and revising.

Connections