Articles

Operating Principles for Applied Models

The project runs on small, hypothesis-driven experiments, original records of work, and continuous forward motion instead of SOTA chasing.

2026-02-28 ยท Working note

The Main Rule

Applied Models only publishes original work from this project.

That means:

  • original experiments
  • original implementations
  • original observations
  • original failures

It does not exist to collect learning notes or summarize someone else's blog, paper, or thread.

The Unit of Progress

The smallest useful unit of progress is:

  1. one specific hypothesis
  2. a handful of key questions
  3. one focused experiment
  4. one honest publication

This keeps the scope realistic and prevents work from stalling in planning.

SOTA Is Not the Requirement

The goal is not to publish only when the result is state of the art.

For an independent researcher, that standard is too expensive as a default operating mode and can freeze progress completely.

What matters more:

  • the experiment is real
  • the method is clear
  • the result is recorded honestly
  • the work compounds over time

Failed Experiments Still Matter

A failed experiment can still be useful if it tells the truth.

An inconclusive run, a weak baseline, or a broken setup can still reveal:

  • a bad assumption
  • a design flaw
  • a blind spot in the model
  • a better next step

The point is not to make every result impressive. The point is to maintain a real research record.