PersonaChat
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
PersonaChat (test)
2.31Persona Consistency
27
PersonaChat turn-level
0.481Spearman Correlation
20
PersonaChat (Human Evaluation)
3.58Fluency
16
PersonaChat 200 random samples (test)
2.36Persona Consistency
11
PersonaChat 1.0 (test)
100Similarity
11
PersonaChat (test)
0.718CS (Coherence Score)
11
PersonaChat original (test)
67.1F1 Score
11
PersonaChat
100Adherence Score
11
PersonaChat (test)
97.7USR RET
10
PersonaChat
41.52Diversity
9
PersonaChat
19.05BLEU-1
8
PersonaChat (full)
7.8Perplexity
6
PersonaChat
3.03QuantiDCE
5
PersonaChat (Forget Set)
100PDLP
4
PersonaChat (test)
80PDLP
4
PersonaChat
97.7USR RET
4
PersonaChat (test)
86.9R@1 (R20 Context)
3
PersonaChat
—Primary metric
0