Cell · cited by the thesis

complex_qwen3.8-27b_survival_final2_reactive_t32_t06

Qwen3.8-27B on the complex floor, 3 episodes. Cited by the thesis in:

Outcome
Wins (boss defeated, alive)3/3 [0.44, 1.00]
Final HP / exact optimum, mean over wins32%
Died0
Final HP, mean22
Seconds per model call131.7
Parse error rate0.0%
Settings
FloorComplex floor
ModelQwen3.8-27B (qwen3.8-27b, thinking)
Planningreactive
Promptcomplex_bfs_strategy_survival.txt
Temperature0.6
Lookahead depth3
Mechanical overridesoff (all engine replacements and assists disabled)
Engine settings (environment, 6)
MOTA_SUPPRESS_HINTS=0
THINKING_MAX_TOKENS=32768
CONVERSATION_HISTORY_TURNS=0
MOTA_ORDER_TWO_SIDED=1
MOTA_FIGHT_KEY_FLOOD=1
LLM_TIMEOUT=1800
Episodesopen one to replay it turn by turn
#outcomefinal HPHP / optstepsmodel callsreplay
1won2232%5120replay
2won2232%5120replay
3won2232%5120replay
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_qwen3.8-27b_survival_final2_reactive_t32_t06, never over the thesis result):

python scripts/run_catalog.py --codes SQCR

Result file: results/reliability/complex_qwen3.8-27b_survival_final2_reactive_t32_t06.json · updated 2026-09-30