Experiments

41 runs, updated 2026-05-28 03:33:28 UTC
Compare runs
single turnmulti turn (default tools)multi turn (all tools)
Level 1l123_single_gptopen →l12_default_gptopen →l12_all_gptopen →
Level 2l123_single_gptopen →l12_default_gptopen →l12_all_gptopen →
Level 3l123_single_gptopen →l3_default_gptopen →l3_all_gptopen →
All runs
fl5_hw_translation_8xh100_default_gpt→fl5_hw_translation_8xh100_single_turn_gpt→grok_smoke_mt_gpt→l1234_popcorn_default_gpt→l123_single_gpt→l12_all_gpt→l12_default_gpt→l3_all_gpt→l3_default_gpt→l5_hw_translation_all_final_gpt→l5_hw_translation_all_full_20260510_gpt→l5_hw_translation_all_full_gpt→l5_hw_translation_all_gpt→l5_hw_translation_default_final_gpt→l5_hw_translation_default_full_20260510_gpt→l5_hw_translation_default_full_20260511_gpt→l5_hw_translation_default_full_gpt→l5_hw_translation_single_gpt→l5_hw_translation_single_turn_final_gpt→l5_hw_translation_single_turn_full_20260510_gpt→l5_hw_translation_single_turn_full_20260511_gpt→l5_hw_translation_single_turn_full_gpt→l5_hw_translation_single_turn_gpt→pop_l1234_st_gpt→pop_l123_all_gpt→pop_l123_all_gpt_stress_20260515_071035_476589996_gpt→pop_l123_all_gpt_stress_gpt→pop_l123_default_gpt→pop_l123_default_gpt_stress_gpt→pop_l123_popcorn2_bench_all_gpt→pop_l123_popcorn2_bench_default_gpt→pop_l123_popcorn2_bench_st_gpt→pop_l123_popcorn2_default_gpt→pop_l123_popcorn2_st_gpt→pop_l123_st_gpt→s1_l5_hw_translation_8xh100_default_gpt→sl5_hw_translation_8xh100_all_gpt→sl5_hw_translation_8xh100_all_reval30m_gpt→sl5_hw_translation_8xh100_all_reval30m_gpt_FIXED_gpt→the_final_pop_l2_popcorn2_all_gpt→the_final_pop_l2_popcorn2_st_gpt→