Studies & Results

Values of Generative Models
AI Responses • Question-wise AI & Country Mean

Study Templates
MTurk Annotation I • MTurk Annotation II

Visualizations

Encoder Models
Zero-shot NLI-based Evaluation (BERT, SpanishBERT, BART, DistilBART, DeBERTA, RoBERTA, XLM-RoBERTA) • Paraphrases of WVS on BERT • PPL-based Evaluation (BERT)

Question Answering Models
UnifiedQA v2.0 (3b-1363200) • UnifiedQA v2.0 (large-1363200) • UnifiedQA v2.0 (large-1251000) • UnifiedQA v1.0 (large)

Generation Models
GPT-3 • BlenderBot v1