complex_qwen3.6-35b_reactive_raw_mechanical
Qwen3.6-35B on the complex floor, 5 episodes. Not cited by the thesis (development or earlier run).
| Wins (boss defeated, alive) | 0/5 [0.00, 0.43] |
| Final HP / exact optimum, mean over wins | – |
| Died | 5 |
| Final HP, mean | -65 |
| Seconds per model call | 14.2 |
| Parse error rate | 0.0% |
| Floor | Complex floor |
| Model | Qwen3.6-35B (qwen3.6-35b-a3b, thinking) |
| Planning | reactive |
| Prompt | complex_bfs_strategy_mechanical.txt |
| Temperature | 0 |
| Lookahead depth | 3 |
| Mechanical overrides | off (all engine replacements and assists disabled) |
| # | outcome | final HP | HP / opt | steps | model calls | replay |
|---|---|---|---|---|---|---|
| 1 | died | -64 | – | 49 | 19 | log not stored |
| 2 | died | -68 | – | 49 | 19 | log not stored |
| 3 | died | -64 | – | 49 | 19 | log not stored |
| 4 | died | -64 | – | 49 | 19 | log not stored |
| 5 | died | -64 | – | 49 | 19 | log not stored |
In a checkout of mt-ai-lab with an API key set (the result goes to a new file rerun_<time>_complex_qwen3.6-35b_reactive_raw_mechanical, never over the thesis result):
python scripts/run_catalog.py --codes A1
Result file: results/reliability/complex_qwen3.6-35b_reactive_raw_mechanical.json · updated 2026-09-22