2WikiMultiHopQA
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
2WikiMultihopQA
70.33F1 Score
9
2WikiMultiHopQA
65.78Accuracy
9
2WikiMultiHopQA v1.0 (test)
2.29Task Latency (s)
9
2WikiMultiHopQA
90.01Refusal F1
8
2WikiMultihopQA in-domain (dev)
94.2Recall@4
8
2WikiMultihopQA
45.7PR@5
8
2WikiMultiHopQA
45,406Retrieved Tokens
8
2WikiMultiHopQA Wikipedia-18 retriever context (test)
36.82EM
7
2WikiMultiHopQA l2
0.09Cost per Pass
7
2WikiMultihopQA (test)
48.01Throughput (Samples/s)
7
2WikiMultihopQA
80.9Recall@3
7
2WikiMultiHopQA
81.53LLM Judge Score
6
2WikiMultiHopQA
69.7F1
6
2WikiMultihopQA 6 (test)
68p-hat
6
2WikiMultiHopQA Out-of-Distribution
70.3Accuracy (8K Context)
6
2WikiMultiHopQA
70.11LM Score
6
2WikiMultiHopQA sampled 1,000 questions (val)
47.4EM (Swedish)
6
2WikiMultiHopQA full-wiki (dev)
28.4MBE
6
2WikiMultiHopQA first 1000 samples (dev)
75EM
6
2WikiMultiHopQA subsampled to 300 Q&A instances
52.07Bridge F1
6
2WikiMultiHopQA n=1,000 samples (test)
70.33F1 Score
5
2WikiMultiHopQA Medium-scale corpus
70.8EM
5
2WikiMultiHopQA
73.4Performance (8K Context)
5
2WikiMultihopQA
0.901Performance
5
2WikiMultiHopQA
72.98Recall@2
4