Question Answering on Average (StrategyQA, MMLU, TruthfulQA, ARC-Challenge)
80.2AUROCDISCOUQ-LLM
Evaluation results
| Method | Links | ||
|---|---|---|---|
| DISCOUQ-LLMID=M12026.03 | 80.2 | 3.6 | |
| LLM Agg.ID=B62026.03 | 79.1 | 9.8 | |
| DISCOUQ-LearnID=M32026.03 | 79 | 4.2 | |
| DISCOUQ-EmbedID=M22026.03 | 77.3 | 4 | |
| Embed Centr.ID=B52026.03 | 77.2 | 16 | |
| Vote CountID=B12026.03 | 77 | 8.4 | |
| Vote EntropyID=B22026.03 | 76.9 | 20.6 | |
| Self-Consist.ID=B42026.03 | 76.9 | 20.6 | |
| Verb. Conf.ID=B32026.03 | 72.4 | 12.8 |