{"tasks":[{"slug":"c6-mode-use-klang","scope":"international","geography":"Klang Valley, Malaysia","phase":4,"family":"individual","scores":[{"slug":"deepseek","value":0.6645985,"skill":-26.72294785012872,"n_valid":200,"name":"DeepSeek-V4-Flash (API, temp 0)","display":"0.665","protocol":"DeepSeek-V4-Flash (API, temp 0)","url":"/tasks/c6-mode-use-klang"},{"slug":"llama","value":0.676809,"skill":-29.051196491562603,"n_valid":200,"name":"Llama-3.3-70B (API, temp 0)","display":"0.677","protocol":"Llama-3.3-70B (API, temp 0)","url":"/tasks/c6-mode-use-klang"},{"slug":"qwen","value":0.682007,"skill":-30.042330060062937,"n_valid":200,"name":"Qwen3.8-27B (API, temp 0)","display":"0.682","protocol":"Qwen3.8-27B (API, temp 0)","url":"/tasks/c6-mode-use-klang"},{"slug":"glm53","value":0.6835450000000001,"skill":-30.335589665363738,"n_valid":200,"name":"GLM-5.3 (API, temp 0)","display":"0.684","protocol":"GLM-5.3 (API, temp 0)","url":"/tasks/c6-mode-use-klang"},{"slug":"minimax3","value":0.6845640000000001,"skill":-30.529888454571473,"n_valid":200,"name":"MiniMax-M3 (API, temp 0)","display":"0.685","protocol":"MiniMax-M3 (API, temp 0)","url":"/tasks/c6-mode-use-klang"},{"slug":"gptoss","value":0.7340300000000001,"skill":-39.96186481075414,"n_valid":200,"name":"gpt-oss-120B (API, temp 0)","display":"0.734","protocol":"gpt-oss-120B (API, temp 0)","url":"/tasks/c6-mode-use-klang"},{"slug":"gemma","value":0.747801,"skill":-42.58766326627899,"n_valid":200,"name":"Gemma-4-26B (API, temp 0)","display":"0.748","protocol":"Gemma-4-26B (API, temp 0)","url":"/tasks/c6-mode-use-klang"}],"n_items":200,"n_test":200,"metric":"Brier (sum over modes)","higher":false,"unit":"","baseline":0.52445,"baseline_display":"0.524","source":"Datasets_TMC, Kah Mun Ng, Universiti Malaya (2021)","url":"/tasks/c6-mode-use-klang","version":"historical","human":null,"best":{"slug":"deepseek","value":0.6645985,"skill":-26.72294785012872,"n_valid":200,"name":"DeepSeek-V4-Flash (API, temp 0)","display":"0.665","protocol":"DeepSeek-V4-Flash (API, temp 0)","url":"/tasks/c6-mode-use-klang"},"leaders":[{"slug":"deepseek","value":0.6645985,"skill":-26.72294785012872,"n_valid":200,"name":"DeepSeek-V4-Flash (API, temp 0)","display":"0.665","protocol":"DeepSeek-V4-Flash (API, temp 0)","url":"/tasks/c6-mode-use-klang"}],"status":"evaluated","title":"Predict which travel modes a person uses, from their profile"}],"aggregation":"Per task and version only; no cross-task average"}