rk-harness findings

Explicit methods scored in Q15 fixed point, and implicit and adaptive methods measured outside the archive.

The rk run ended on 2026-11-10, and no more compute goes to it. Epoch 2 is frozen and these pages are its final snapshot: the site is no longer rebuilt. The overview site sums up the final findings, and the harness is public for anyone to download and run.

An automated search for Runge-Kutta methods that hold up in Q15 fixed-point arithmetic on small microcontrollers. Cost is counted against a modeled Cortex-M0+ cycle budget rather than measured on the chip. This site was rebuilt from the run's data at the end of every cycle until the run ended, and nobody edits it by hand. The project overview explains why the project exists and how it works.

Built from cycle 17,364 of the run, 358,752 archive records. Newest elite recorded 2026-10-23 23:10 CT, after the last saturation check at 2026-09-27 17:23 CT.

The three classes

A scored record has room for one explicit tableau and nothing else, so the implicit and adaptive classes are measured outside the archive and never ranked against it.

explicit
358,752
archive records, from the run archive

Scored, archived and ranked at an equal cycle budget.

Q15 fixed point, checked by the pinned verifier.

explicit methods

implicit
47
measured ledger points, from the measurement ledger

Measured because the explicit class runs out of stability before it runs out of budget.

float64; not order-verified and not scored.

implicit methods

adaptive
56
measured ledger points, from the measurement ledger

A record has no room for the second b vector an error estimate needs.

float64, with one fixed pair also run in Q15; not order-verified and not scored.

adaptive methods

The three classes side by side

explicitimplicitadaptive
method familyExplicit fixed-step Runge-Kutta: a strictly lower triangular A and one b vectorTwo-stage SDIRK over a dyadic grid of gamma and a21, with a Newton iteration count and a Jacobian policyAn embedded pair on the dyadic lattice (order 3 with an order-2 estimate) and a step-size controller
how candidates are foundMAP-Elites cells filled by exhaustive enumeration and CMA-ES islandsDeterministic enumeration, plus the side-track jobsDeterministic enumeration, plus the side-track jobs
the question askedWhat error does a method reach inside a fixed cycle budget?How many modeled cycles does a method need to reach an accuracy target? (cycles to tolerance)How many modeled cycles does a method need to reach an accuracy target? (cycles to tolerance)
problemsThe frozen search set and held-out setThe validation problemsThe validation problems
arithmeticQ15 through the pinned solverfloat64: there is no Q15 LU factorizationfloat64; the Q15 adaptive solver runs one fixed pair
what checks a resultThe pinned verifier, order conditions included, then scoring on the held-out setNothing: not order-verified and not scoredNothing: not order-verified and not scored
cost basisThe pinned cost modelA design estimate that assumes a division costFloat64 attempt counts, each attempt priced by the Q15 cost model
source documentsrk-work/archive, validation/results.json, benchmark/results.jsonimplicit_archive/elites.json, the side-track ledger, validation/results.json, benchmark/results.jsonadaptive_archive/elites.json, the side-track ledger, benchmark/results.json

Numbers from different columns do not compare, so each class tab carries its own.

Epoch status

Epoch 2 frozen scored method class: explicit fixed-step Runge-Kutta

frozen at
2026-11-10 13:35 CT
reason
The owner ended the run on 2026-11-10. No further cycles will run, so these numbers will not change.
epoch 1
frozen copy of the epoch 1 site, frozen 2026-09-14 23:33 CT
last progress
2026-10-23 23:10 CT: a cell elite improved its held-out error (elite_improvement)
saturation counter
0 consecutive saturating checks; 6 trigger a freeze
last check
2026-09-27 17:23 CT, verdict CONTINUE
falsification file
not yet produced

Progress is a record in an empty cell, an improved elite or a heldout_verified acceptance. With no progress inside the saturation window and the falsification file written, a check counts as saturating, and enough in a row reach the freeze threshold. When the next epoch opens, the frozen epoch's pages are copied to a directory of their own, and the copy keeps the scoring hash they were built under. Starting the next epoch, and re-pinning the verifier for it, is a person's decision. Times are stored in UTC and shown in US Central.