Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

Response Generation research benchmarks – Page 2 · SOTA2 Research

Response Generation

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

ReDial v1 (test)

28BLEU

6

Feb 26, 2026

Empathetic Dialogue

MCTS-Driven Knowledge Retrieval

75Pairwise Diversity

6

Feb 26, 2026

MCTS-Driven Knowledge Retrieval

78.5Pairwise Diversity

6

Feb 26, 2026

ChattyChef (test)

5.4BLEU

6

Feb 26, 2026

96.4Match Acc

6

Feb 26, 2026

57.09F1 Score

5

Jun 4, 2026

MultiDoc2Dial (test)

SF-Re2Gw/o subgraph

43.65F1 Score

5

Jun 4, 2026

13.73BLEU-4

5

May 26, 2026

ReDial (Human Evaluation)

2.43Relevance Score

5

Feb 26, 2026

13.84Perplexity (PPL)

5

Feb 26, 2026

50BLEU-1

5

Feb 26, 2026

0.866Match Score

5

Feb 26, 2026

Audio Visual Scene-aware Dialog (AVSD) v1.0 (val)

0.851CIDEr

4

May 26, 2026

GoRecDial 1.0 (test)

29.6BLEU

4

Feb 26, 2026

University Admission Inquiries (test)

Fine-Tuned with RAG

92.7Fact Recall

4

Feb 26, 2026

23.6BLEU-1

4

Feb 26, 2026

85.1CIDEr

4

Feb 26, 2026

RolePlayConv (evaluation set)

ModeratorLM-Think

7.4Response Score

3

Jun 12, 2026

HuggingGPT Human Evaluation Set 130 diverse requests (test)

63.08Success Rate

3

Feb 26, 2026

Chain-of-Thought

61Win Count

3

Feb 26, 2026

Chain-of-Thought

0.95Win Rate

3

Feb 26, 2026

0.005BLEU

2

Feb 26, 2026

Human Evaluation multi-turn (200 items) (test)

72Adversarial Win Rate

1

Feb 26, 2026

Human Evaluation (200 items) single-turn (test)

62Adversarial Win Rate

1

Feb 26, 2026

Page 2 of 2

Popular datasets

REDIAL Human Evaluation DailyDialog ESConv AdvBench

Follow for updates

Join SOTA2 Contact