Cell

segment2_qwen3.5-397b_basic_d1d4

Qwen3.5-397B on the tower (MT8→MT11), 2 runs. Not cited by the thesis (development or earlier run).

Outcome
Highest floor reached, per runMT8, MT8
Died0
Final HP, mean823
Seconds per model call–
Parse error rate–
Settings
FloorTower MT8→MT11
ModelQwen3.5-397B (qwen3.5-397b-a17b, thinking)
Planningreactive
Promptengine-default-tower-basic
Temperature0
Lookahead depth2
Mechanical overrideson (logged per turn)
Walkthrough checkpoint2
Engine settings (environment, 1)
THINKING_MAX_TOKENS=8192
Runsopen one to replay it turn by turn
#outcomefinal HPHP / optturnsmodel callsreplay
1MT8823–10replay
2MT8823–10replay
Re-run this cell

In a checkout of mt-ai-lab with an API key set (the result goes to a new file rerun_<time>_segment2_qwen3.5-397b_basic_d1d4, never over the thesis result):

python scripts/run_catalog.py --codes W2

Result file: results/reliability/segment2_qwen3.5-397b_basic_d1d4.json · updated 2026-09-26