Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

EduFeedback

Benchmarks

Task NameDataset NameSOTA ResultTrend
Multi-turn Conversation EvaluationEduFeedback alternate
MT-Bench Score8.3
20
LLM AlignmentEduFeedback
TFLOPs152.56
20
Preference AlignmentEduFeedback-Alternate (test)
Pairwise Win Rate (excl. ties)67.78
5
Showing 3 of 3 rows