{"label":"gs_llama8b_full","preset":"llama-3.1-8b","task_model":"meta-llama-3.1-8b-instruct","task_mode":"instruct","obs":"full","train":["simple","gen01","gen02"],"test":["gen03","gen04","gen05","gen06","gen07","gen08","gen09","gen10","gen11","gen12","gen13","gen14","gen15"],"repeat":1,"env":{"CONVERSATION_HISTORY_TURNS":"0","MOTA_ORDER_TWO_SIDED":"1","MOTA_FIGHT_KEY_FLOOD":"1","LOOKAHEAD_DEPTH":"3","MOTA_BARE_OBS":null,"MOTA_EVAL_TEMPERATURE":"0.0","THINKING_MAX_TOKENS":"16384","MOTA_DISABLE_OVERRIDES":null,"MOTA_SUPPRESS_HINTS":null},"overrides":"mechanical-only (dead-end / unwinnable / lethal) + auto-collect + auto-fight-0-HP","episodes":[{"prompt":"core","floor":"simple","split":"train","rep":0,"goal_score":0.325,"success":false,"hp_ratio":null,"died":false,"final_hp":410,"hero_atk":13,"hero_def":8,"enemies_killed":13,"enemies_total":14,"llm_calls":15,"parse_errors":0,"forced_turns":21,"lethal_only_stop":null,"run_log":"20260928_003616_373365_simple_llm.json","elapsed_s":55},{"prompt":"gepa","floor":"simple","split":"train","rep":0,"goal_score":0.325,"success":false,"hp_ratio":null,"died":false,"final_hp":410,"hero_atk":13,"hero_def":8,"enemies_killed":13,"enemies_total":14,"llm_calls":15,"parse_errors":0,"forced_turns":21,"lethal_only_stop":null,"run_log":"20260928_003711_363882_simple_llm.json","elapsed_s":55.9},{"prompt":"basic","floor":"simple","split":"train","rep":0,"goal_score":0.325,"success":false,"hp_ratio":null,"died":false,"final_hp":424,"hero_atk":13,"hero_def":8,"enemies_killed":13,"enemies_total":14,"llm_calls":16,"parse_errors":2,"forced_turns":20,"lethal_only_stop":null,"run_log":"20260928_003807_314919_simple_llm.json","elapsed_s":77},{"prompt":"core","floor":"gen01","split":"train","rep":0,"goal_score":0.7768,"success":true,"hp_ratio":0.5536,"died":false,"final_hp":93,"hero_atk":14,"hero_def":7,"enemies_killed":10,"enemies_total":10,"llm_calls":8,"parse_errors":0,"forced_turns":17,"lethal_only_stop":null,"run_log":"20260928_003924_332106_gen01.txt_llm.json","elapsed_s":26.6},{"prompt":"gepa","floor":"gen01","split":"train","rep":0,"goal_score":0.7768,"success":true,"hp_ratio":0.5536,"died":false,"final_hp":93,"hero_atk":14,"hero_def":7,"enemies_killed":10,"enemies_total":10,"llm_calls":8,"parse_errors":0,"forced_turns":17,"lethal_only_stop":null,"run_log":"20260928_003950_923095_gen01.txt_llm.json","elapsed_s":26.9},{"prompt":"basic","floor":"gen01","split":"train","rep":0,"goal_score":0.7619,"success":true,"hp_ratio":0.5238,"died":false,"final_hp":88,"hero_atk":14,"hero_def":7,"enemies_killed":10,"enemies_total":10,"llm_calls":8,"parse_errors":0,"forced_turns":17,"lethal_only_stop":null,"run_log":"20260928_004017_789655_gen01.txt_llm.json","elapsed_s":34.4},{"prompt":"core","floor":"gen02","split":"train","rep":0,"goal_score":0.9052,"success":true,"hp_ratio":0.8103,"died":false,"final_hp":188,"hero_atk":14,"hero_def":7,"enemies_killed":9,"enemies_total":12,"llm_calls":9,"parse_errors":0,"forced_turns":14,"lethal_only_stop":null,"run_log":"20260928_004052_214849_gen02.txt_llm.json","elapsed_s":49.8},{"prompt":"gepa","floor":"gen02","split":"train","rep":0,"goal_score":0.9052,"success":true,"hp_ratio":0.8103,"died":false,"final_hp":188,"hero_atk":14,"hero_def":7,"enemies_killed":9,"enemies_total":12,"llm_calls":9,"parse_errors":0,"forced_turns":14,"lethal_only_stop":null,"run_log":"20260928_004142_124521_gen02.txt_llm.json","elapsed_s":41.4},{"prompt":"basic","floor":"gen02","split":"train","rep":0,"goal_score":0.612,"success":true,"hp_ratio":0.2241,"died":false,"final_hp":52,"hero_atk":14,"hero_def":7,"enemies_killed":12,"enemies_total":12,"llm_calls":11,"parse_errors":2,"forced_turns":15,"lethal_only_stop":null,"run_log":"20260928_004223_518015_gen02.txt_llm.json","elapsed_s":83.9},{"prompt":"core","floor":"gen03","split":"test","rep":0,"goal_score":0.2864,"success":false,"hp_ratio":null,"died":false,"final_hp":501,"hero_atk":13,"hero_def":6,"enemies_killed":9,"enemies_total":11,"llm_calls":9,"parse_errors":0,"forced_turns":11,"lethal_only_stop":null,"run_log":"20260928_004347_429169_gen03.txt_llm.json","elapsed_s":33},{"prompt":"gepa","floor":"gen03","split":"test","rep":0,"goal_score":0.2864,"success":false,"hp_ratio":null,"died":false,"final_hp":510,"hero_atk":13,"hero_def":6,"enemies_killed":9,"enemies_total":11,"llm_calls":10,"parse_errors":0,"forced_turns":10,"lethal_only_stop":null,"run_log":"20260928_004420_412940_gen03.txt_llm.json","elapsed_s":36.4},{"prompt":"basic","floor":"gen03","split":"test","rep":0,"goal_score":0.2864,"success":false,"hp_ratio":null,"died":false,"final_hp":510,"hero_atk":13,"hero_def":6,"enemies_killed":9,"enemies_total":11,"llm_calls":10,"parse_errors":1,"forced_turns":10,"lethal_only_stop":null,"run_log":"20260928_004456_858519_gen03.txt_llm.json","elapsed_s":49.4},{"prompt":"core","floor":"gen04","split":"test","rep":0,"goal_score":0.72,"success":true,"hp_ratio":0.44,"died":false,"final_hp":99,"hero_atk":14,"hero_def":8,"enemies_killed":11,"enemies_total":11,"llm_calls":9,"parse_errors":0,"forced_turns":20,"lethal_only_stop":null,"run_log":"20260928_004546_242425_gen04.txt_llm.json","elapsed_s":32.9},{"prompt":"gepa","floor":"gen04","split":"test","rep":0,"goal_score":0.72,"success":true,"hp_ratio":0.44,"died":false,"final_hp":99,"hero_atk":14,"hero_def":8,"enemies_killed":11,"enemies_total":11,"llm_calls":9,"parse_errors":0,"forced_turns":20,"lethal_only_stop":null,"run_log":"20260928_004619_183876_gen04.txt_llm.json","elapsed_s":32.2},{"prompt":"basic","floor":"gen04","split":"test","rep":0,"goal_score":0.7045,"success":true,"hp_ratio":0.4089,"died":false,"final_hp":92,"hero_atk":14,"hero_def":8,"enemies_killed":11,"enemies_total":11,"llm_calls":8,"parse_errors":0,"forced_turns":21,"lethal_only_stop":null,"run_log":"20260928_004651_381030_gen04.txt_llm.json","elapsed_s":33},{"prompt":"core","floor":"gen05","split":"test","rep":0,"goal_score":0.5843,"success":true,"hp_ratio":0.1686,"died":false,"final_hp":29,"hero_atk":14,"hero_def":8,"enemies_killed":11,"enemies_total":13,"llm_calls":12,"parse_errors":0,"forced_turns":14,"lethal_only_stop":null,"run_log":"20260928_004724_346983_gen05.txt_llm.json","elapsed_s":45.5},{"prompt":"gepa","floor":"gen05","split":"test","rep":0,"goal_score":0.7791,"success":true,"hp_ratio":0.5581,"died":false,"final_hp":96,"hero_atk":14,"hero_def":8,"enemies_killed":10,"enemies_total":13,"llm_calls":11,"parse_errors":0,"forced_turns":14,"lethal_only_stop":null,"run_log":"20260928_004809_893414_gen05.txt_llm.json","elapsed_s":42.7},{"prompt":"basic","floor":"gen05","split":"test","rep":0,"goal_score":0.3231,"success":false,"hp_ratio":null,"died":false,"final_hp":234,"hero_atk":14,"hero_def":8,"enemies_killed":12,"enemies_total":13,"llm_calls":13,"parse_errors":1,"forced_turns":14,"lethal_only_stop":null,"run_log":"20260928_004852_605630_gen05.txt_llm.json","elapsed_s":64.2},{"prompt":"core","floor":"gen06","split":"test","rep":0,"goal_score":0.8347,"success":true,"hp_ratio":0.6694,"died":false,"final_hp":162,"hero_atk":14,"hero_def":7,"enemies_killed":5,"enemies_total":8,"llm_calls":4,"parse_errors":0,"forced_turns":15,"lethal_only_stop":null,"run_log":"20260928_004956_760659_gen06.txt_llm.json","elapsed_s":13.5},{"prompt":"gepa","floor":"gen06","split":"test","rep":0,"goal_score":0.8347,"success":true,"hp_ratio":0.6694,"died":false,"final_hp":162,"hero_atk":14,"hero_def":7,"enemies_killed":5,"enemies_total":8,"llm_calls":4,"parse_errors":0,"forced_turns":15,"lethal_only_stop":null,"run_log":"20260928_005010_225712_gen06.txt_llm.json","elapsed_s":14},{"prompt":"basic","floor":"gen06","split":"test","rep":0,"goal_score":1,"success":true,"hp_ratio":1,"died":false,"final_hp":242,"hero_atk":14,"hero_def":7,"enemies_killed":4,"enemies_total":8,"llm_calls":3,"parse_errors":1,"forced_turns":15,"lethal_only_stop":null,"run_log":"20260928_005024_208252_gen06.txt_llm.json","elapsed_s":15.4},{"prompt":"core","floor":"gen07","split":"test","rep":0,"goal_score":0.8716,"success":true,"hp_ratio":0.7432,"died":false,"final_hp":165,"hero_atk":14,"hero_def":9,"enemies_killed":14,"enemies_total":14,"llm_calls":13,"parse_errors":0,"forced_turns":18,"lethal_only_stop":null,"run_log":"20260928_005039_623243_gen07.txt_llm.json","elapsed_s":52},{"prompt":"gepa","floor":"gen07","split":"test","rep":0,"goal_score":0.8716,"success":true,"hp_ratio":0.7432,"died":false,"final_hp":165,"hero_atk":14,"hero_def":9,"enemies_killed":14,"enemies_total":14,"llm_calls":13,"parse_errors":0,"forced_turns":18,"lethal_only_stop":null,"run_log":"20260928_005131_627943_gen07.txt_llm.json","elapsed_s":51.4},{"prompt":"basic","floor":"gen07","split":"test","rep":0,"goal_score":0.9166,"success":true,"hp_ratio":0.8333,"died":false,"final_hp":185,"hero_atk":14,"hero_def":9,"enemies_killed":13,"enemies_total":14,"llm_calls":11,"parse_errors":1,"forced_turns":19,"lethal_only_stop":null,"run_log":"20260928_005223_016635_gen07.txt_llm.json","elapsed_s":74.5},{"prompt":"core","floor":"gen08","split":"test","rep":0,"goal_score":0.704,"success":true,"hp_ratio":0.408,"died":false,"final_hp":102,"hero_atk":14,"hero_def":8,"enemies_killed":8,"enemies_total":11,"llm_calls":9,"parse_errors":0,"forced_turns":16,"lethal_only_stop":null,"run_log":"20260928_005337_544068_gen08.txt_llm.json","elapsed_s":39.1},{"prompt":"gepa","floor":"gen08","split":"test","rep":0,"goal_score":0.744,"success":true,"hp_ratio":0.488,"died":false,"final_hp":122,"hero_atk":14,"hero_def":8,"enemies_killed":8,"enemies_total":11,"llm_calls":9,"parse_errors":0,"forced_turns":16,"lethal_only_stop":null,"run_log":"20260928_005416_674319_gen08.txt_llm.json","elapsed_s":39.6},{"prompt":"basic","floor":"gen08","split":"test","rep":0,"goal_score":0.744,"success":true,"hp_ratio":0.488,"died":false,"final_hp":122,"hero_atk":14,"hero_def":8,"enemies_killed":8,"enemies_total":11,"llm_calls":9,"parse_errors":2,"forced_turns":16,"lethal_only_stop":null,"run_log":"20260928_005456_241070_gen08.txt_llm.json","elapsed_s":53.3},{"prompt":"core","floor":"gen09","split":"test","rep":0,"goal_score":0.6421,"success":true,"hp_ratio":0.2843,"died":false,"final_hp":58,"hero_atk":14,"hero_def":8,"enemies_killed":11,"enemies_total":11,"llm_calls":11,"parse_errors":1,"forced_turns":18,"lethal_only_stop":null,"run_log":"20260928_005549_495963_gen09.txt_llm.json","elapsed_s":41.1},{"prompt":"gepa","floor":"gen09","split":"test","rep":0,"goal_score":0.6397,"success":true,"hp_ratio":0.2794,"died":false,"final_hp":57,"hero_atk":14,"hero_def":8,"enemies_killed":11,"enemies_total":11,"llm_calls":10,"parse_errors":0,"forced_turns":19,"lethal_only_stop":null,"run_log":"20260928_005630_588235_gen09.txt_llm.json","elapsed_s":37.7},{"prompt":"basic","floor":"gen09","split":"test","rep":0,"goal_score":0.8652,"success":true,"hp_ratio":0.7304,"died":false,"final_hp":149,"hero_atk":14,"hero_def":8,"enemies_killed":10,"enemies_total":11,"llm_calls":9,"parse_errors":1,"forced_turns":19,"lethal_only_stop":null,"run_log":"20260928_005708_243863_gen09.txt_llm.json","elapsed_s":43.1},{"prompt":"core","floor":"gen10","split":"test","rep":0,"goal_score":0.2917,"success":false,"hp_ratio":null,"died":false,"final_hp":311,"hero_atk":14,"hero_def":6,"enemies_killed":10,"enemies_total":12,"llm_calls":9,"parse_errors":0,"forced_turns":15,"lethal_only_stop":null,"run_log":"20260928_005751_381920_gen10.txt_llm.json","elapsed_s":36.7},{"prompt":"gepa","floor":"gen10","split":"test","rep":0,"goal_score":0.2917,"success":false,"hp_ratio":null,"died":false,"final_hp":311,"hero_atk":14,"hero_def":6,"enemies_killed":10,"enemies_total":12,"llm_calls":9,"parse_errors":1,"forced_turns":15,"lethal_only_stop":null,"run_log":"20260928_005828_069954_gen10.txt_llm.json","elapsed_s":38.8},{"prompt":"basic","floor":"gen10","split":"test","rep":0,"goal_score":0.2917,"success":false,"hp_ratio":null,"died":false,"final_hp":300,"hero_atk":14,"hero_def":6,"enemies_killed":10,"enemies_total":12,"llm_calls":9,"parse_errors":3,"forced_turns":15,"lethal_only_stop":null,"run_log":"20260928_005906_828226_gen10.txt_llm.json","elapsed_s":49.6},{"prompt":"core","floor":"gen11","split":"test","rep":0,"goal_score":0.1944,"success":false,"hp_ratio":null,"died":false,"final_hp":459,"hero_atk":13,"hero_def":7,"enemies_killed":5,"enemies_total":9,"llm_calls":8,"parse_errors":0,"forced_turns":13,"lethal_only_stop":null,"run_log":"20260928_005956_406121_gen11.txt_llm.json","elapsed_s":29.7},{"prompt":"gepa","floor":"gen11","split":"test","rep":0,"goal_score":0.1944,"success":false,"hp_ratio":null,"died":false,"final_hp":414,"hero_atk":13,"hero_def":7,"enemies_killed":5,"enemies_total":9,"llm_calls":8,"parse_errors":0,"forced_turns":13,"lethal_only_stop":null,"run_log":"20260928_010026_148456_gen11.txt_llm.json","elapsed_s":30.6},{"prompt":"basic","floor":"gen11","split":"test","rep":0,"goal_score":0.1944,"success":false,"hp_ratio":null,"died":false,"final_hp":468,"hero_atk":13,"hero_def":7,"enemies_killed":5,"enemies_total":9,"llm_calls":7,"parse_errors":0,"forced_turns":14,"lethal_only_stop":null,"run_log":"20260928_010056_772596_gen11.txt_llm.json","elapsed_s":33.1},{"prompt":"core","floor":"gen12","split":"test","rep":0,"goal_score":0.6456,"success":true,"hp_ratio":0.2912,"died":false,"final_hp":53,"hero_atk":13,"hero_def":4,"enemies_killed":3,"enemies_total":11,"llm_calls":3,"parse_errors":1,"forced_turns":10,"lethal_only_stop":null,"run_log":"20260928_010129_894049_gen12.txt_llm.json","elapsed_s":14.2},{"prompt":"gepa","floor":"gen12","split":"test","rep":0,"goal_score":0.6456,"success":true,"hp_ratio":0.2912,"died":false,"final_hp":53,"hero_atk":13,"hero_def":4,"enemies_killed":3,"enemies_total":11,"llm_calls":3,"parse_errors":0,"forced_turns":10,"lethal_only_stop":null,"run_log":"20260928_010144_049090_gen12.txt_llm.json","elapsed_s":13.8},{"prompt":"basic","floor":"gen12","split":"test","rep":0,"goal_score":1,"success":true,"hp_ratio":1,"died":false,"final_hp":182,"hero_atk":13,"hero_def":4,"enemies_killed":2,"enemies_total":11,"llm_calls":3,"parse_errors":0,"forced_turns":10,"lethal_only_stop":null,"run_log":"20260928_010157_839507_gen12.txt_llm.json","elapsed_s":14.1},{"prompt":"core","floor":"gen13","split":"test","rep":0,"goal_score":0.3208,"success":false,"hp_ratio":null,"died":false,"final_hp":257,"hero_atk":14,"hero_def":9,"enemies_killed":11,"enemies_total":12,"llm_calls":9,"parse_errors":0,"forced_turns":17,"lethal_only_stop":{"action":"Fight Boss at (5,1) dmg=264","hp_after":-7,"damage":264},"run_log":"20260928_010211_930716_gen13.txt_llm.json","elapsed_s":37.9},{"prompt":"gepa","floor":"gen13","split":"test","rep":0,"goal_score":0.7995,"success":true,"hp_ratio":0.599,"died":false,"final_hp":124,"hero_atk":14,"hero_def":9,"enemies_killed":9,"enemies_total":12,"llm_calls":7,"parse_errors":0,"forced_turns":17,"lethal_only_stop":null,"run_log":"20260928_010249_897301_gen13.txt_llm.json","elapsed_s":28.4},{"prompt":"basic","floor":"gen13","split":"test","rep":0,"goal_score":0.6594,"success":true,"hp_ratio":0.3188,"died":false,"final_hp":66,"hero_atk":14,"hero_def":9,"enemies_killed":11,"enemies_total":12,"llm_calls":9,"parse_errors":0,"forced_turns":17,"lethal_only_stop":null,"run_log":"20260928_010318_345647_gen13.txt_llm.json","elapsed_s":44.4},{"prompt":"core","floor":"gen14","split":"test","rep":0,"goal_score":0.3111,"success":false,"hp_ratio":null,"died":false,"final_hp":474,"hero_atk":13,"hero_def":9,"enemies_killed":8,"enemies_total":9,"llm_calls":7,"parse_errors":0,"forced_turns":15,"lethal_only_stop":null,"run_log":"20260928_010402_745491_gen14.txt_llm.json","elapsed_s":25.6},{"prompt":"gepa","floor":"gen14","split":"test","rep":0,"goal_score":0.7616,"success":true,"hp_ratio":0.5232,"died":false,"final_hp":79,"hero_atk":13,"hero_def":9,"enemies_killed":7,"enemies_total":9,"llm_calls":6,"parse_errors":0,"forced_turns":15,"lethal_only_stop":null,"run_log":"20260928_010428_335224_gen14.txt_llm.json","elapsed_s":21.2},{"prompt":"basic","floor":"gen14","split":"test","rep":0,"goal_score":0.8377,"success":true,"hp_ratio":0.6755,"died":false,"final_hp":102,"hero_atk":13,"hero_def":9,"enemies_killed":6,"enemies_total":9,"llm_calls":5,"parse_errors":1,"forced_turns":15,"lethal_only_stop":null,"run_log":"20260928_010449_522063_gen14.txt_llm.json","elapsed_s":24.7},{"prompt":"core","floor":"gen15","split":"test","rep":0,"goal_score":0.824,"success":true,"hp_ratio":0.6481,"died":false,"final_hp":140,"hero_atk":14,"hero_def":9,"enemies_killed":10,"enemies_total":12,"llm_calls":8,"parse_errors":0,"forced_turns":20,"lethal_only_stop":null,"run_log":"20260928_010514_244349_gen15.txt_llm.json","elapsed_s":31.8},{"prompt":"gepa","floor":"gen15","split":"test","rep":0,"goal_score":0.824,"success":true,"hp_ratio":0.6481,"died":false,"final_hp":140,"hero_atk":14,"hero_def":9,"enemies_killed":10,"enemies_total":12,"llm_calls":8,"parse_errors":0,"forced_turns":20,"lethal_only_stop":null,"run_log":"20260928_010546_053609_gen15.txt_llm.json","elapsed_s":30},{"prompt":"basic","floor":"gen15","split":"test","rep":0,"goal_score":0.7592,"success":true,"hp_ratio":0.5185,"died":false,"final_hp":112,"hero_atk":14,"hero_def":9,"enemies_killed":11,"enemies_total":12,"llm_calls":9,"parse_errors":1,"forced_turns":20,"lethal_only_stop":null,"run_log":"20260928_010616_074836_gen15.txt_llm.json","elapsed_s":43.6}],"started":"2026-09-28T02:04:26.838044","gepa":{"result_file":"results/gepa/gepa_result_gs_llama8b_full.json","best_score":0.669,"trajectory":[0.669,0.5619,0.6654,0.575,0.5799,0.5619],"accepted":[true,true,true,true,true,null],"best_is_seed":true,"finished":"2026-09-28T02:36:16.345395"},"updated":"2026-09-28T03:06:59.667695","summary":{"train/core":{"n":3,"success":2,"deaths":0,"goal_mean":0.669},"train/gepa":{"n":3,"success":2,"deaths":0,"goal_mean":0.669},"train/basic":{"n":3,"success":2,"deaths":0,"goal_mean":0.5663},"test/core":{"n":13,"success":8,"deaths":0,"goal_mean":0.5562},"test/gepa":{"n":13,"success":10,"deaths":0,"goal_mean":0.6456},"test/basic":{"n":13,"success":9,"deaths":0,"goal_mean":0.6602},"all/core":{"n":16,"success":10,"deaths":0,"goal_mean":0.5774},"all/gepa":{"n":16,"success":12,"deaths":0,"goal_mean":0.65},"all/basic":{"n":16,"success":11,"deaths":0,"goal_mean":0.6426}},"finished":"2026-09-28T03:06:59.667695"}