Cognitive FabExpertise › polyx
Expertise · journals become rules

polyx — rules mined from
your own history

A rule that has held across your last two hundred runs is a lookup, not an inference. polyx is the engine for that layer: it turns the journals your work already produces into candidate rules, scores them against a corpus before any of them is adopted, and keeps the evidence attached so a rule can be read, challenged, and re-scored when the corpus moves.

Status: polyx is early. This page describes the mechanism, not a product you can install today. There is no distribution, no package and no repository to link — if this is the layer you need, say so and it will shape what ships first.

The shape of the mechanism

1 · Journals in

The input is history you already have: runs, the decisions taken inside them, and what came of those decisions. Nothing has to be written from memory, and no model is asked to invent a rule base.

2 · Candidates out

Recurring structure in that history becomes a set of candidate rules — each one an explicit statement about what holds, not a summary of what usually happens.

3 · Scored against a corpus

A candidate is scored before it is allowed to carry weight. Support is a number over real history: how much of the corpus the rule accounts for, and where it fails.

4 · Adopted with its evidence

An adopted rule travels with the runs behind it. That is what lets somebody who was not there read it, and somebody who disagrees with it argue about the evidence rather than the phrasing.

5 · Re-scored when the corpus moves

History keeps changing. A rule that held and stops holding is caught by the mechanism that admitted it. Learning here is replay, not retraining.

6 · Looked up, not inferred

At run time the rule set is consulted, not reasoned about. Applying it costs nothing, gives the same answer every time, and is the thing you point at when someone asks why.

Why it belongs beside the weights, not inside them

Adding expertise this way means adding artifacts, not a training run — so it can be adopted, audited and reversed on the timescale of a code review. A rule that stops matching the corpus is retired without anything being unlearned, and an agent that meets a genuinely new situation still has the whole model available for it.

It is also the only form in which the deterministic part of the work becomes inspectable. A step left to inference is a step you cannot check before it runs, cannot replay afterwards, and cannot explain to anyone who asks why. A scored rule is all three.