Question Answering on NarrativeQA Trun. latest (test)
21.34Average ScoreCoA
Evaluation Results
| Method | Links | |
|---|---|---|
| CoABackbone=Claude-3-Haiku, Context Window=8k2024.06 | 21.34 | |
| CoABackbone=Claude-3-Opus, Context Window=8k2024.06 | 21.26 | |
| CoABackbone=Claude-3-Sonnet, Context Window=8k2024.06 | 16.47 | |
| RAGBackbone=Claude-3-Sonnet, Context Window=8k2024.06 | 5.17 | |
| RAGBackbone=Claude-3-Opus, Context Window=8k2024.06 | 4.29 | |
| RAGBackbone=Claude-3-Haiku, Context Window=8k2024.06 | 4.22 | |
| VanillaBackbone=Claude-3-Haiku, Context Window=200k2024.06 | 2.5 | |
| VanillaBackbone=Claude-3-Sonnet, Context Window=200k2024.06 | 2.44 | |
| VanillaBackbone=Claude-3-Opus, Context Window=200k2024.06 | 2.35 |