Execution attempt

Sign in with GitHub
← Experiment E2 · Does APT4's tokenization of digits and number separators differ from the Mistral-derived tokenizer's, and does the difference account for the lower GSM8K score of Bielik-PL-11B-v3.0-Instruct and for accuracy differences between English and Polish number formats?

Execution attempt · A15 · planned

Grades the probe generations and computes per-cell accuracy, the pre-registered contrasts and carry gradients.

Succeeded · Registered by @stw2 via agent. Reported and received times are kept apart; no computation or result is verified.

Pinned source and configuration

https://github.com/stw2/tokenizer-science-tax @ c43f19850803969bc1fefe537caf14e54cfd9a9f

Reference checked 2026-09-14 13:08 UTC. Later commits, branches or plan changes do not retarget this attempt.

Command
python scripts/07_grade.py
Working directory
experiments/E02-digit-policy
Configuration paths
None
Parameters
31 golden tests; paired item bootstrap of 1,000 resamples, seed 20260703; Holm over C1, C4', C2 for the PL model (space to no-break space) and C5; strata with both-model accuracy between 0.05 and 0.95; format contrasts on items of at least four digits
Environment
CPython 3.12.13 (uv 0.11.16); CPU; Apple silicon, 128 GB unified memory
Output directory
Not recorded

Inputs

Materials the registrant named when registering this attempt, by content identity; obtainability is derived from their location reports. Nothing is fetched or verified.

Delivered events

  1. Registered

    #1

    Grades the probe generations and computes per-cell accuracy, the pre-registered contrasts and carry gradients.

    reported · received · @stw2 via agent · posted to the Thread

  2. Started

    #2

    Started.

    reported · received · @stw2 via agent · attempt only

  3. Succeeded

    #3

    Completed with exit code 0.

    Exit code 0.

    • graded_bielik-pl-11b-v3.0-instruct.jsonl · https://github.com/stw2/tokenizer-science-tax/blob/2fc1a1b415a06b7648cdb37319ad573ccdc96a81/experiments/E02-digit-policy/results/graded_bielik-pl-11b-v3.0-instruct.jsonl · public · sha256 07f65d753e3f… · 5469197 bytes · M83
    • graded_bielik-11b-v3.0-instruct.jsonl · https://github.com/stw2/tokenizer-science-tax/blob/2fc1a1b415a06b7648cdb37319ad573ccdc96a81/experiments/E02-digit-policy/results/graded_bielik-11b-v3.0-instruct.jsonl · public · sha256 c5b60083c014… · 5410788 bytes · M84
    • probe_results.json · https://github.com/stw2/tokenizer-science-tax/blob/2fc1a1b415a06b7648cdb37319ad573ccdc96a81/experiments/E02-digit-policy/results/probe_results.json · public · sha256 b64683edf833… · 22513 bytes · M85

    reported · received · @stw2 via agent · posted to the Thread

Report an event

The registrant’s capture tool normally delivers start and outcome events. Reporting here is the same author report with the browser as the reported time; it does not observe the process.

This attempt has a delivered outcome. A new execution is a new attempt.