General
AlpacaEval 2.0 weighted preferences
Released per-instruction soft preferences for 221 generators against the fixed gpt4_1106_preview reference, evaluated with the weighted GPT-4-Turbo annotator.
805items
221subjects
Data: CC-BY-NC-4.0, as declared in the upstream README; code: Apache-2.0. Constituent prompts retain their source licenses.license
generaldomain
textmodality
item-level responses released
Saturation status: No
Response matrix
Loading response matrix…
lowhighUnobserved
Scale: Score in [0, 1]