List your company

ResearchFormerly WizWand

PubMedQA

Benchmarks

TaskDatasetSOTA resultTrendResults

PubMedQA 1,000 questions, 1,000-abstract corpus

0.98MRR

6

Question Answering

PubMedQA

66.4Acc

6

Question Answering

PubMedQA

9.7BLEU-1

6

Graph-based Agent Memory Poisoning

PubMedQA

100Utilization (Util.)

5

Hallucination Detection

PubMedQA

94.8AUROC

5

Question Answering

PubMedQA Retrieval OS Setting (test)

76.4Accuracy

5

Question Answering

PubMedQA

73Accuracy

5

Question Answering

PubmedQA

43.21F1

5

Question Answering

PubMedQA English (test)

74.64Accuracy

5

Medical Question Answering

PubMedQA

77.6Accuracy

4

Medical Reasoning

PubMedQA (test)

58.3Accuracy

4

Question Answering

PubMedQA Zero-Context Setting (test)

64.77Accuracy

4

Medical Question Answering

PubMedQA (test)

12.2CUS Score

4

Question Answering

PubMedQA official (val)

93.33F1 Score

4

PubMedQA (test)

37.49ROUGE-1

4

Question Answering

PubMedQA 80/20 split (seed 42) (eval)

0Pathwise Violations

3

Factual Grounding Evaluation

PubmedQA

0.84ROC-AUC

3

Question Answering

PubMedQA

800Average Token Consumption

3

Question Answering

PubMedQA

75Accuracy

2

Biomedical Question Answering

PubMedQA

0.6Baseline Accuracy

2

Biomedical Question Answering

PubMedQA (test)

0.254CUS

2

Medical Question Answering

PubMedQA

66CUS Score

2

Explanation Quality Evaluation

PubMedQA (test)

39.7Reasoning Soundness Loss

2

Knowledge Graph Question Answering

PubMedQA adaptive (test)

27Base Error

1

Question Answering

PubMedQA

0Selection Rate

1