Cell

hand01_deepseek-v4-flash_selflookahead

DeepSeek-V4-Flash on hand01, 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, mean76
Seconds per model call43.5
Parse error rate3.4%
Settings
Floorhand01
ModelDeepSeek-V4-Flash (deepseek-v4-flash-0731, thinking)
Planningreactive
Promptself_lookahead.txt
Temperature0
Lookahead depth3
Mechanical overrideson (logged per turn)
Engine settings (environment, 7)
THINKING_MAX_TOKENS=32768
CONVERSATION_HISTORY_TURNS=0
MOTA_ORDER_TWO_SIDED=1
MOTA_FIGHT_KEY_FLOOD=1
MOTA_BARE_OBS=1
MOTA_GRID_MAP=1
LLM_TIMEOUT=1800
Episodesopen one to replay it turn by turn
#outcomefinal HPHP / optstepsmodel callsreplay
1not won133–3216replay
2not won33–179replay
3not won48–178replay
4not won33–179replay
5not won133–3217replay
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>_hand01_deepseek-v4-flash_selflookahead, never over the thesis result):

python scripts/run_catalog.py --codes SLH

Result file: results/reliability/hand01_deepseek-v4-flash_selflookahead.json · updated 2026-10-02