Cell

complex_qwen3.5-397b_final2_la3_t06

Qwen3.5-397B on the complex floor, 5 episodes. Not cited by the thesis (development or earlier run).

Outcome
Wins (boss defeated, alive)0/5 [0.00, 0.43]
Final HP / exact optimum, mean over wins–
Died0
Final HP, mean73
Seconds per model call180.4
Parse error rate2.6%
Settings
FloorComplex floor
ModelQwen3.5-397B (qwen3.5-397b-a17b, thinking)
Planningreactive
Promptengine-default-floor-basic
Temperature0.6
Lookahead depth3
Mechanical overrideson (logged per turn)
Engine settings (environment, 5)
THINKING_MAX_TOKENS=32768
CONVERSATION_HISTORY_TURNS=0
MOTA_ORDER_TWO_SIDED=1
MOTA_FIGHT_KEY_FLOOD=1
LLM_TIMEOUT=300
Episodesopen one to replay it turn by turn
#outcomefinal HPHP / optstepsmodel callsreplay
1not won120–3715replay
2not won34–3916replay
3not won84–3715replay
4not won66–3715replay
5not won61–4217replay
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>_complex_qwen3.5-397b_final2_la3_t06, never over the thesis result):

python scripts/run_catalog.py --codes T6GC

Result file: results/reliability/complex_qwen3.5-397b_final2_la3_t06.json · updated 2026-09-30