Datasets:
id stringlengths 19 19 | text stringlengths 757 1.86k | label stringclasses 3
values |
|---|---|---|
mednli_train_000001 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000002 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000003 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000004 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000005 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000006 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000007 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000008 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000009 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000010 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000011 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000012 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000013 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000014 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000015 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000016 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000017 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000018 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000019 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000020 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000021 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000022 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000023 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000024 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000025 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000026 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000027 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000028 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000029 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000030 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000031 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000032 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000033 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000034 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000035 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000036 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000037 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000038 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000039 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000040 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000041 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000042 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000043 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000044 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000045 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000046 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000047 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000048 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000049 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000050 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000051 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000052 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000053 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000054 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000055 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000056 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000057 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000058 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000059 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000060 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000061 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000062 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000063 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000064 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000065 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000066 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000067 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000068 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000069 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000070 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000071 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000072 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000073 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000074 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000075 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000076 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000077 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000078 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000079 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000080 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000081 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000082 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000083 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000084 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000085 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000086 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000087 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000088 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000089 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000090 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000091 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000092 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000093 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000094 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000095 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000096 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000097 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
mednli_train_000098 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | CONTRADICTION |
mednli_train_000099 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | NEUTRAL |
mednli_train_000100 | You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not state... | ENTAILMENT |
MedNLI with Instructions
Dataset Summary
MedNLI with Instructions is a prompt-formatted version of the MedNLI three-class natural language inference (NLI) benchmark.
Each example combines the original clinical premise and hypothesis with explicit natural-language instructions, inference-direction rules, label definitions, and a constrained output format.
This transformation is designed to support research on instruction-based classification and generalization from general-domain NLI to clinical-domain NLI.
The original examples and gold labels are preserved. The test split is shuffled without sampling.
Dataset Structure
Each example contains three fields:
| Column | Description |
|---|---|
id |
Unique identifier for each example |
text |
Premise and hypothesis combined with classification instructions |
label |
Original NLI label converted to uppercase |
Labels
The dataset uses three classification labels:
- ENTAILMENT: The hypothesis is definitely true given the premise.
- NEUTRAL: The hypothesis might be true or false given the premise.
- CONTRADICTION: The hypothesis is definitely false given the premise.
Instruction Format
Each example follows the format below:
You are a system for natural language inference classification.
Input:
Two marked spans:
- <p>...</p> = PREMISE
- <h>...</h> = HYPOTHESIS
Rules:
- Determine the relationship of the HYPOTHESIS with respect to the PREMISE.
- Use only the information provided in the PREMISE.
- Do NOT assume information that is not stated or logically implied.
- Do NOT reverse the direction of inference.
Labels:
- ENTAILMENT = The hypothesis is definitely true given the premise.
- NEUTRAL = The hypothesis might be true or false given the premise.
- CONTRADICTION = The hypothesis is definitely false given the premise.
Input Pair:
<p> {premise} </p>
<h> {hypothesis} </h>
Output EXACTLY one label: ENTAILMENT, CONTRADICTION, or NEUTRAL.
DO NOT output anything else.
Here, {premise} and {hypothesis} are replaced with the corresponding original MedNLI sentences.
Dataset Splits
| Split | Number of Examples |
|---|---|
| Train | 11,232 |
| Validation | 1,395 |
| Test | 1,422 |
| Total | 14,049 |
All original examples are retained.
The test examples are shuffled using random seed 42. No sampling or deduplication is performed.
Data Preprocessing
The original MedNLI data contains the following columns:
sentence1: Premisesentence2: Hypothesisgold_label: Ground-truth NLI label
The preprocessing pipeline performs the following operations:
- Reads the original train, validation, and test splits.
- Extracts the premise from
sentence1. - Extracts the hypothesis from
sentence2. - Inserts the premise and hypothesis into
<p> ... </p>and<h> ... </h>tags. - Adds explicit NLI classification instructions.
- Converts labels to uppercase without changing their original meaning.
- Assigns a unique ID to every example.
- Shuffles the entire test split using random seed
42. - Saves the processed examples as JSONL files.
The processed files contain:
train.jsonl
validation.jsonl
test.jsonl
Unique Identifiers
Each example receives a unique identifier.
Examples:
mednli_train_000001
mednli_validation_000001
mednli_test_000001
The identifiers are generated during preprocessing and are not the original MedNLI pair IDs.
Research Applications
This dataset format is intended for research on:
- Natural language inference classification
- Instruction-based fine-tuning
- Clinical-domain generalization
- Cross-domain transfer from SNLI to MedNLI
- Robustness and generalization of language models
- Comparisons between instruction-based and sentence-only training strategies
The instruction-based and sentence-only versions use identical examples, labels, and IDs, enabling controlled comparisons.
Source Dataset
This dataset is derived from MedNLI, introduced by Romanov and Shivade (2018).
MedNLI is a clinical natural language inference dataset constructed from clinical records associated with MIMIC-III.
Original paper:
Lessons from Natural Language Inference in the Clinical Domain (EMNLP 2018)
Official dataset:
Source representation:
presencesw/mednli on Hugging Face
License and Data Access
The original MedNLI dataset is subject to the PhysioNet Credentialed Health Data License and associated access requirements.
Reformatting the original clinical sentences does not eliminate these restrictions.
Redistribution or upload of the underlying examples to Hugging Face or another third-party hosting service requires appropriate authorization.
This dataset card describes the preprocessing procedure and does not itself grant redistribution rights.
See:
- Downloads last month
- 96