Table Selection on SPIDER 2018 (test)
2.7Avg TablesReAct
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| ReActSelector=Qwen-2.5-7B-Instruct, Dense Retrieval=UAE-Large-V12026.01 | 2.7 | 67.6 | 96.7 | 75.5 | 93.4 | |
| CORE-TSelector=Qwen-2.5-7B-Instruct, Dense Retrieval=UAE-Large-V12026.01 | 2.9 | 54.3 | 97 | 66.8 | 94.5 | |
| CORE-TSelector=Llama-3.1-8B-Instruct, Dense Retrieval=UAE-Large-V12026.01 | 4.2 | 40.3 | 98 | 53.8 | 96 | |
| ReActSelector=Llama-3.1-8B-Instruct, Dense Retrieval=UAE-Large-V12026.01 | 4.5 | 39.7 | 96.4 | 52.2 | 92.8 | |
| DR@5Selector=Llama-3.1-8B-Instruct, Dense Retrieval=UAE-Large-V12026.01 | 5 | 29.4 | 98 | 43.9 | 95.6 | |
| DRR@5Selector=Llama-3.1-8B-Instruct, Dense Retrieval=UAE-Large-V12026.01 | 5 | 30 | 99.5 | 44.8 | 98.7 | |
| JAR@5Selector=Llama-3.1-8B-Instruct, Dense Retrieval=UAE-Large-V12026.01 | 5 | 29.6 | 98.5 | 44.2 | 96.7 | |
| DR@5Selector=Qwen-2.5-7B-Instruct, Dense Retrieval=UAE-Large-V12026.01 | 5 | 29.6 | 98.3 | 44.1 | 96.4 | |
| DRR@5Selector=Qwen-2.5-7B-Instruct, Dense Retrieval=UAE-Large-V12026.01 | 5 | 30 | 99.5 | 44.8 | 98.8 | |
| JAR@5Selector=Qwen-2.5-7B-Instruct, Dense Retrieval=UAE-Large-V12026.01 | 5 | 29.5 | 98.3 | 44.1 | 96.3 |