GPT 4o Mini (2024-07-18)

problems=279 · window_start=2024-08-01

problems=454 · window_start=2024-08-01

arena=text · category=coding · style_control=true

arena=text · category=industry_legal_and_government · style_control=true

arena=text · category=industry_software_and_it_services · style_control=true

arena=text · category=french · style_control=true

arena=text · category=industry_medicine_and_healthcare · style_control=true

arena=text · category=english · style_control=true

arena=text · category=industry_life_and_physical_and_social_science · style_control=true

arena=text · category=longer_query · style_control=true

arena=text · category=hard_prompts_english · style_control=true

arena=text · category=overall · style_control=true

arena=text · category=chinese · style_control=true

arena=text · category=industry_writing_and_literature_and_language · style_control=true

arena=text · category=multi_turn · style_control=true

arena=text · category=hard_prompts · style_control=true

arena=text · category=industry_business_and_management_and_financial_operations · style_control=true

arena=text · category=spanish · style_control=true

arena=text · category=russian · style_control=true

arena=text · category=german · style_control=true

arena=text · category=non_english · style_control=true

arena=text · category=creative_writing · style_control=true

arena=text · category=instruction_following · style_control=true

arena=text · category=industry_mathematical · style_control=true

arena=text · category=industry_entertainment_and_sports_and_media · style_control=true

arena=text · category=expert · style_control=true

arena=text · category=math · style_control=true

arena=text · category=exclude_ties · style_control=true

arena=text · category=japanese · style_control=true

arena=text · category=korean · style_control=true

implementation=vals-ai

implementation=vals-ai

implementation=vals-ai

implementation=vals-ai

split=semi_private · model_type=Base LLM

split=public_eval · model_type=Base LLM

split=public_eval · model_type=Base LLM

split=semi_private · model_type=Base LLM

implementation=vals-ai