Question Answering on NarrativeQA No Trun. latest (test)
23.96Average ScoreCoA
Evaluation Results
| Method | Links | |
|---|---|---|
| CoABackbone=Claude-3-Opus, Context Window=8k2024.06 | 23.96 | |
| CoABackbone=Claude-3-Haiku, Context Window=8k2024.06 | 18.8 | |
| CoABackbone=Claude-3-Sonnet, Context Window=8k2024.06 | 16.51 | |
| RAGBackbone=Claude-3-Sonnet, Context Window=8k2024.06 | 8.5 | |
| VanillaBackbone=Claude-3-Haiku, Context Window=200k2024.06 | 7.17 | |
| VanillaBackbone=Claude-3-Opus, Context Window=200k2024.06 | 6.56 | |
| RAGBackbone=Claude-3-Opus, Context Window=8k2024.06 | 5.86 | |
| RAGBackbone=Claude-3-Haiku, Context Window=8k2024.06 | 5.71 | |
| VanillaBackbone=Claude-3-Sonnet, Context Window=200k2024.06 | 5.15 | |
| CoABackbone=Claude-3-Opus, Context Window=8k2024.06 | 0.2438 | |
| CoABackbone=Claude-3-Haiku, Context Window=8k2024.06 | 0.1831 | |
| CoABackbone=Claude-3-Sonnet, Context Window=8k2024.06 | 0.1663 | |
| RAGBackbone=Claude-3-Sonnet, Context Window=8k2024.06 | 0.0909 | |
| VanillaBackbone=Claude-3-Haiku, Context Window=200k2024.06 | 0.08 | |
| VanillaBackbone=Claude-3-Opus, Context Window=200k2024.06 | 0.0723 | |
| RAGBackbone=Claude-3-Opus, Context Window=8k2024.06 | 0.0613 | |
| RAGBackbone=Claude-3-Haiku, Context Window=8k2024.06 | 0.0594 | |
| VanillaBackbone=Claude-3-Sonnet, Context Window=200k2024.06 | 0.0558 |