GENERAL RULES FOR READING THE EXAMPLES

  1. A "bad indicator" is a proxy that makes it possible to obtain a reward without the required result.
  2. The function, indicator, standard and calculation rule are established before the task is performed.
  3. Specific control cases and answers may be hidden and must be rotated if the public set can be memorised.
  4. The "verified actual value" is taken from a trusted source or a reproducible check to which the evaluated policy cannot write the final result.
  5. A self-report, a log created by the agent itself and a repeated model output do not become independent verified actual values merely by being stored in a database.
  6. An artefact under verification is treated as untrusted data and cannot give commands to the evaluator.
  7. "Quality" is not mixed with the quantity of the result; mandatory self-checks relate to quality rather than to additional costs.
  8. Overperformance on one indicator must not cover a critical failure on another.
  9. If the result cannot yet be measured, this is treated as a technology bug, and no variable reward is credited until the technology is corrected. However, this impossibility must be justified in writing: it must state exactly what prevents measurement, which measurement methods were considered and rejected and when the issue will be reviewed.
  10. A finite set of counterexamples is the beginning of red-team testing, not proof that no new exploits exist.
  11. Late confirmation may change the verified actual value of the original attempt under the previous version of the rules; this is not a retrospective change to the benchmark.
  12. A discovered rewarded exploit is tested for transfer and may require quarantine of the data and policy.

About the author

Sergei Sychev is a member of the TRIZ-RI Group research community (Israel, Slovakia, Czech Republic), and has been an expert in the Theory of Inventive Problem Solving (TRIZ) since 1985.

The author would prefer that the reader's attention be focused on the material itself, rather than on the author's details, as would be appropriate when reading scientific papers. If you would like to get in touch, write to or reach out on.