Execution attempt

Sign in with GitHub
← Experiment E10 · Does the end-to-end success gap between Bielik-11B-v3.0-Instruct and Bielik-PL-11B-v3.0-Instruct on multi-step tool-use tasks exceed the gap predicted from their one-shot per-step success rates?

Execution attempt · A103 · rerun

Runs all 100 instances with English and Polish instructions through the tool loop with Bielik-11B-v3.0-Instruct.

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 @ d35f2683bfcbd5dfb12264c6b8880e7a783eb192

Reference checked 2026-09-14 13:17 UTC. Later commits, branches or plan changes do not retarget this attempt.

Command
python scripts/03_run_traj.py --model orig
Working directory
experiments/E10-agentic-compounding
Configuration paths
None
Parameters
model orig; languages en,pl; at most 8 steps; 2048 generated tokens per step; context window 32768; observations cut at 2000 characters; 20 s per step; greedy
Environment
3.12.13 (uv 0.11.16); MLX bf16 on Metal, greedy decoding (temperature 0); 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

    Runs all 100 instances with English and Polish instructions through the tool loop with Bielik-11B-v3.0-Instruct.

    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.

    • traj_orig.jsonl · Not redistributed (observations contain home-directory paths; the public copy replaces them): ask the Room owner, or rebuild with scripts/03_run_traj.py at the pinned revisions · restricted · sha256 f25cc44a8204… · 773903 bytes · M474
    • traj_orig.jsonl, home paths replaced · https://github.com/stw2/tokenizer-science-tax/blob/76ed9b8efec6102935361df4b013764d057e4363/experiments/E10-agentic-compounding/results/raw/traj_orig.jsonl · public · sha256 8b36b7ff7f2c… · 759943 bytes · M475

    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.