Execution attempt

Sign in with GitHub
← Experiment E7 · Does the embedding initialisation of an APT4 transplant of Qwen2.5-1.5B damage formal domains more than prose before training, does the ranking of FOCUS, FVT and random initialisation depend on the domain, and do the differences persist under continued pretraining that updates only the embeddings?

Execution attempt · A75 · planned

Computes the stage-2 endpoints: random-minus-FVT gaps on formal domains, FVT−FOCUS persistence, recovery curves, tokens to 90% recovery and the embeddings-only share of full-parameter recovery.

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 @ 82a236f723c0f3191bfe3b4bc44002fbdd6e80c4

Reference checked 2026-09-14 13:12 UTC. Later commits, branches or plan changes do not retarget this attempt.

Command
python scripts/09_analyze_stage2.py
Working directory
experiments/E07-embedding-init
Configuration paths
None
Parameters
bootstrap resamples 1000; seed 20260705; endpoint checkpoint 150M; run time not recorded, reported at the results commit
Environment
3.12.13; numpy on 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

    Computes the stage-2 endpoints: random-minus-FVT gaps on formal domains, FVT−FOCUS persistence, recovery curves, tokens to 90% recovery and the embeddings-only share of full-parameter recovery.

    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.

    • analysis_s2.json · https://github.com/stw2/tokenizer-science-tax/blob/7d450d1f62209ff5ac5415543443d2a25db74b7e/experiments/E07-embedding-init/results/analysis_s2.json · public · sha256 2c2b9afec737… · 8711 bytes · M392

    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.