Cell

simple_qwen3.8-27b_reactive

Qwen3.8-27B on the simple floor, 5 episodes. Not cited by the thesis (development or earlier run).

Episodesopen one to replay it turn by turn

#	outcome	final HP	HP / opt	steps	model calls	replay
1	won	2	15%	36	–	log not stored
2	won	2	15%	36	–	log not stored
3	not won	464	–	36	–	log not stored
4	won	2	15%	36	–	log not stored
5	won	2	15%	36	–	log not stored

Wins (boss defeated, alive)	4/5 [0.38, 0.96]
Final HP / exact optimum, mean over wins	15%
Died	0
Final HP, mean	94
Seconds per model call	–
Parse error rate	–