Single-hop Question Answering on NarrativeQA
23.89ScoreDCS
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| DCSBackbone=Llama-3-8B-Instruct2025.06 | 23.89 | — | |
| Llama-3-8B-InstructBackbone=Llama-3-8B-Instruct, Context Management=Baseline2025.06 | 21.54 | — | |
| Mistral-7B-InstructBackbone=Mistral-7B-Instruct, Context Management=Baseline2025.06 | 20.89 | — | |
| InfLLMBackbone=Llama-3-8B-Instruct2025.06 | 19.62 | — | |
| StreamingBackbone=Llama-3-8B-Instruct2025.06 | 19.3 | — | |
| LM-infiniteBackbone=Llama-3-8B-Instruct2025.06 | 18.83 | — | |
| DCSBackbone=Mistral-7B-Instruct2025.06 | 18.63 | — | |
| MOICEBackbone=Mistral-7B-Instruct2025.06 | 17.03 | — | |
| InfLLMBackbone=Vicuna-7B2025.06 | 16.07 | — | |
| DCSBackbone=Vicuna-7B2025.06 | 15.6 | — | |
| StreamingBackbone=Vicuna-7B2025.06 | 15.37 | — | |
| Vicuna-7BBackbone=Vicuna-7B, Context Management=Baseline2025.06 | 14.95 | — | |
| MOICEBackbone=Vicuna-7B2025.06 | 14.84 | — | |
| InfLLMBackbone=Mistral-7B-Instruct2025.06 | 14.59 | — | |
| LM-InfiniteBackbone=Vicuna-7B2025.06 | 14.12 | — | |
| LM-infiniteBackbone=Mistral-7B-Instruct2025.06 | 12.87 | — | |
| StreamingBackbone=Mistral-7B-Instruct2025.06 | 12.6 | — | |
| Llama-3-8B-Instruct with DCSBackbone=Llama-3-8B-Instruct, Context Strategy=DCS2025.06 | 0.2389 | 0.355 | |
| Llama-3-8B-InstructBackbone=Llama-3-8B-Instruct, Context Strategy=None2025.06 | 0.2154 | 0.276 | |
| Llama-3-8B-Instruct with InfilmBackbone=Llama-3-8B-Instruct, Context Strategy=Infilm2025.06 | 0.1962 | 0.2715 | |
| Llama-3-8B-Instruct with StreamingBackbone=Llama-3-8B-Instruct, Context Strategy=Streaming2025.06 | 0.193 | 0.2426 | |
| Llama-3-8B-Instruct with LM-infiniteBackbone=Llama-3-8B-Instruct, Context Strategy=LM-infinite2025.06 | 0.1883 | 0.242 |