messages listlengths 3 3 |
|---|
[
{
"role": "system",
"content": "You are a scientific claim extractor. Extract claims preserving ALL qualifiers (may, suggests, not significant, preliminary, etc). Output JSON array of claims with: text, epistemic_tag, confidence, qualifiers, is_null_result, status, missing_fields."
},
{
"role": "use... |
[
{
"role": "system",
"content": "You are the Decision Agent of a PhD Research OS. Given the current research goals, knowledge gaps, and incomplete/low-confidence claims, propose a Decision Object.\n\nA Decision Object includes:\n- decision_id: string (DEC_XXXX)\n- recommended_action: one of \"experiment\", \... |
[
{
"role": "system",
"content": "You are the Query Planner of a PhD Research OS. Given a broad research question, decompose it into 2-4 specific sub-queries that can be independently searched in a scientific knowledge base.\n\nEach sub-query should:\n- Target a specific aspect of the question\n- Be answerabl... |
[
{
"role": "system",
"content": "You are the Verifier Agent of a PhD Research OS. Given two scientific claims, determine if they contradict each other.\n\nAnalyze the claims and output a Conflict Resolution Object:\n- conflict_detected: boolean\n- conflict_type: one of \"value_mismatch\", \"methodology_diffe... |
[
{
"role": "system",
"content": "You are the Researcher Agent of a PhD Research OS. Your role is to extract structured scientific claims from research paper text.\n\nFor each claim, output a JSON object with these fields:\n- claim_id: string (CLM_XXXX format)\n- text: the claim text as stated in the paper\n-... |
[
{
"role": "system",
"content": "You are the Query Planner of a PhD Research OS. Given a broad research question, decompose it into 2-4 specific sub-queries that can be independently searched in a scientific knowledge base.\n\nEach sub-query should:\n- Target a specific aspect of the question\n- Be answerabl... |
[
{
"role": "system",
"content": "You are the Confidence Scorer of a PhD Research OS. Score the confidence of a scientific claim using this formula:\n\nconfidence = evidence_strength × study_quality_weight × journal_tier_weight × completeness_penalty\n\nWhere:\n- evidence_strength [0,1]: How directly the evid... |
[
{
"role": "system",
"content": "You are the Query Planner of a PhD Research OS. Given a broad research question, decompose it into 2-4 specific sub-queries that can be independently searched in a scientific knowledge base.\n\nEach sub-query should:\n- Target a specific aspect of the question\n- Be answerabl... |
[
{
"role": "system",
"content": "You are a hallucination detector for a PhD Research OS. Given source text and an extracted claim, determine if the claim is faithful to the source. Output JSON: {\"faithful\": boolean, \"error_type\": \"none|hallucinated_value|negation_flipped|qualifier_stripped|scope_overrea... |
[
{
"role": "system",
"content": "You are the Researcher Agent of a PhD Research OS. Your role is to extract structured scientific claims from research paper text.\n\nFor each claim, output a JSON object with these fields:\n- claim_id: string (CLM_XXXX format)\n- text: the claim text as stated in the paper\n-... |
[
{
"role": "system",
"content": "You are the Epistemic Classifier of a PhD Research OS. Given a scientific statement, classify it into exactly one category:\n\n- Fact: Directly supported by experimental data with quantitative evidence. Reproducible measurements.\n- Interpretation: Author's explanation of dat... |
[
{
"role": "system",
"content": "You are the Query Planner of a PhD Research OS. Given a broad research question, decompose it into 2-4 specific sub-queries that can be independently searched in a scientific knowledge base.\n\nEach sub-query should:\n- Target a specific aspect of the question\n- Be answerabl... |
[
{
"role": "system",
"content": "You are the Researcher Agent of a PhD Research OS. Your role is to extract structured scientific claims from research paper text.\n\nFor each claim, output a JSON object with these fields:\n- claim_id: string (CLM_XXXX format)\n- text: the claim text as stated in the paper\n-... |
[
{
"role": "system",
"content": "You are the Researcher Agent of a PhD Research OS. Your role is to extract structured scientific claims from research paper text.\n\nFor each claim, output a JSON object with these fields:\n- claim_id: string (CLM_XXXX format)\n- text: the claim text as stated in the paper\n-... |
[
{
"role": "system",
"content": "You are the Decision Agent of a PhD Research OS. Given the current research goals, knowledge gaps, and incomplete/low-confidence claims, propose a Decision Object.\n\nA Decision Object includes:\n- decision_id: string (DEC_XXXX)\n- recommended_action: one of \"experiment\", \... |
[
{
"role": "system",
"content": "You are the Query Planner of a PhD Research OS. Given a broad research question, decompose it into 2-4 specific sub-queries that can be independently searched in a scientific knowledge base.\n\nEach sub-query should:\n- Target a specific aspect of the question\n- Be answerabl... |
[
{
"role": "system",
"content": "You are a scientific claim extractor. Extract claims preserving ALL qualifiers (may, suggests, not significant, preliminary, etc). Output JSON array of claims with: text, epistemic_tag, confidence, qualifiers, is_null_result, status, missing_fields."
},
{
"role": "use... |
[
{
"role": "system",
"content": "You are the Verifier Agent of a PhD Research OS. Given two scientific claims, determine if they contradict each other.\n\nAnalyze the claims and output a Conflict Resolution Object:\n- conflict_detected: boolean\n- conflict_type: one of \"value_mismatch\", \"methodology_diffe... |
[
{
"role": "system",
"content": "You are the Epistemic Classifier of a PhD Research OS. Given a scientific statement, classify it into exactly one category:\n\n- Fact: Directly supported by experimental data with quantitative evidence. Reproducible measurements.\n- Interpretation: Author's explanation of dat... |
[
{
"role": "system",
"content": "You are the Confidence Scorer of a PhD Research OS. Score the confidence of a scientific claim using this formula:\n\nconfidence = evidence_strength × study_quality_weight × journal_tier_weight × completeness_penalty\n\nWhere:\n- evidence_strength [0,1]: How directly the evid... |
[
{
"role": "system",
"content": "You are the Researcher Agent of a PhD Research OS. Your role is to extract structured scientific claims from research paper text.\n\nFor each claim, output a JSON object with these fields:\n- claim_id: string (CLM_XXXX format)\n- text: the claim text as stated in the paper\n-... |
[
{
"role": "system",
"content": "You are the Confidence Scorer of a PhD Research OS. Score the confidence of a scientific claim using this formula:\n\nconfidence = evidence_strength × study_quality_weight × journal_tier_weight × completeness_penalty\n\nWhere:\n- evidence_strength [0,1]: How directly the evid... |
[
{
"role": "system",
"content": "You are the Confidence Scorer of a PhD Research OS. Score the confidence of a scientific claim using this formula:\n\nconfidence = evidence_strength × study_quality_weight × journal_tier_weight × completeness_penalty\n\nWhere:\n- evidence_strength [0,1]: How directly the evid... |
[
{
"role": "system",
"content": "You are the Epistemic Classifier of a PhD Research OS. Given a scientific statement, classify it into exactly one category:\n\n- Fact: Directly supported by experimental data with quantitative evidence. Reproducible measurements.\n- Interpretation: Author's explanation of dat... |
[
{
"role": "system",
"content": "You are the Researcher Agent of a PhD Research OS. Your role is to extract structured scientific claims from research paper text.\n\nFor each claim, output a JSON object with these fields:\n- claim_id: string (CLM_XXXX format)\n- text: the claim text as stated in the paper\n-... |
[
{
"role": "system",
"content": "You are a scientific claim extractor. Extract claims preserving ALL qualifiers (may, suggests, not significant, preliminary, etc). Output JSON array of claims with: text, epistemic_tag, confidence, qualifiers, is_null_result, status, missing_fields."
},
{
"role": "use... |
[
{
"role": "system",
"content": "You are the Epistemic Classifier of a PhD Research OS. Given a scientific statement, classify it into exactly one category:\n\n- Fact: Directly supported by experimental data with quantitative evidence. Reproducible measurements.\n- Interpretation: Author's explanation of dat... |
[
{
"role": "system",
"content": "You are the Researcher Agent of a PhD Research OS. Your role is to extract structured scientific claims from research paper text.\n\nFor each claim, output a JSON object with these fields:\n- claim_id: string (CLM_XXXX format)\n- text: the claim text as stated in the paper\n-... |
[
{
"role": "system",
"content": "You are the Confidence Scorer of a PhD Research OS. Score the confidence of a scientific claim using this formula:\n\nconfidence = evidence_strength × study_quality_weight × journal_tier_weight × completeness_penalty\n\nWhere:\n- evidence_strength [0,1]: How directly the evid... |
[
{
"role": "system",
"content": "You are a scientific claim extractor. Extract claims preserving ALL qualifiers (may, suggests, not significant, preliminary, etc). Output JSON array of claims with: text, epistemic_tag, confidence, qualifiers, is_null_result, status, missing_fields."
},
{
"role": "use... |
[
{
"role": "system",
"content": "You are the Verifier Agent of a PhD Research OS. Given two scientific claims, determine if they contradict each other.\n\nAnalyze the claims and output a Conflict Resolution Object:\n- conflict_detected: boolean\n- conflict_type: one of \"value_mismatch\", \"methodology_diffe... |
[
{
"role": "system",
"content": "You are the Query Planner of a PhD Research OS. Given a broad research question, decompose it into 2-4 specific sub-queries that can be independently searched in a scientific knowledge base.\n\nEach sub-query should:\n- Target a specific aspect of the question\n- Be answerabl... |
[
{
"role": "system",
"content": "You are the Decision Agent of a PhD Research OS. Given the current research goals, knowledge gaps, and incomplete/low-confidence claims, propose a Decision Object.\n\nA Decision Object includes:\n- decision_id: string (DEC_XXXX)\n- recommended_action: one of \"experiment\", \... |
[
{
"role": "system",
"content": "You are a hallucination detector for a PhD Research OS. Given source text and an extracted claim, determine if the claim is faithful to the source. Output JSON: {\"faithful\": boolean, \"error_type\": \"none|hallucinated_value|negation_flipped|qualifier_stripped|scope_overrea... |
[
{
"role": "system",
"content": "You are the Verifier Agent of a PhD Research OS. Given two scientific claims, determine if they contradict each other.\n\nAnalyze the claims and output a Conflict Resolution Object:\n- conflict_detected: boolean\n- conflict_type: one of \"value_mismatch\", \"methodology_diffe... |
[
{
"role": "system",
"content": "You are the Epistemic Classifier of a PhD Research OS. Given a scientific statement, classify it into exactly one category:\n\n- Fact: Directly supported by experimental data with quantitative evidence. Reproducible measurements.\n- Interpretation: Author's explanation of dat... |
[
{
"role": "system",
"content": "You are a scientific claim extractor. Extract claims preserving ALL qualifiers (may, suggests, not significant, preliminary, etc). Output JSON array of claims with: text, epistemic_tag, confidence, qualifiers, is_null_result, status, missing_fields."
},
{
"role": "use... |
[
{
"role": "system",
"content": "You are the Epistemic Classifier of a PhD Research OS. Given a scientific statement, classify it into exactly one category:\n\n- Fact: Directly supported by experimental data with quantitative evidence. Reproducible measurements.\n- Interpretation: Author's explanation of dat... |
[
{
"role": "system",
"content": "You are the Confidence Scorer of a PhD Research OS. Score the confidence of a scientific claim using this formula:\n\nconfidence = evidence_strength × study_quality_weight × journal_tier_weight × completeness_penalty\n\nWhere:\n- evidence_strength [0,1]: How directly the evid... |
[
{
"role": "system",
"content": "You are a scientific claim extractor. Extract claims preserving ALL qualifiers (may, suggests, not significant, preliminary, etc). Output JSON array of claims with: text, epistemic_tag, confidence, qualifiers, is_null_result, status, missing_fields."
},
{
"role": "use... |
[
{
"role": "system",
"content": "You are the Epistemic Classifier of a PhD Research OS. Given a scientific statement, classify it into exactly one category:\n\n- Fact: Directly supported by experimental data with quantitative evidence. Reproducible measurements.\n- Interpretation: Author's explanation of dat... |
[
{
"role": "system",
"content": "You are the Epistemic Classifier of a PhD Research OS. Given a scientific statement, classify it into exactly one category:\n\n- Fact: Directly supported by experimental data with quantitative evidence. Reproducible measurements.\n- Interpretation: Author's explanation of dat... |
[
{
"role": "system",
"content": "You are the Confidence Scorer of a PhD Research OS. Score the confidence of a scientific claim using this formula:\n\nconfidence = evidence_strength × study_quality_weight × journal_tier_weight × completeness_penalty\n\nWhere:\n- evidence_strength [0,1]: How directly the evid... |
[
{
"role": "system",
"content": "You are extracting claims from the RESULTS section of a scientific paper. Apply appropriate epistemic standards for this section. Abstract claims cannot be tagged as Fact. Discussion claims should be tagged as Interpretation or Hypothesis."
},
{
"role": "user",
"c... |
[
{
"role": "system",
"content": "You are the Researcher Agent of a PhD Research OS. Your role is to extract structured scientific claims from research paper text.\n\nFor each claim, output a JSON object with these fields:\n- claim_id: string (CLM_XXXX format)\n- text: the claim text as stated in the paper\n-... |
[
{
"role": "system",
"content": "You are the Epistemic Classifier of a PhD Research OS. Given a scientific statement, classify it into exactly one category:\n\n- Fact: Directly supported by experimental data with quantitative evidence. Reproducible measurements.\n- Interpretation: Author's explanation of dat... |
[
{
"role": "system",
"content": "You are the Researcher Agent of a PhD Research OS. Your role is to extract structured scientific claims from research paper text.\n\nFor each claim, output a JSON object with these fields:\n- claim_id: string (CLM_XXXX format)\n- text: the claim text as stated in the paper\n-... |
[
{
"role": "system",
"content": "You are the Query Planner of a PhD Research OS. Given a broad research question, decompose it into 2-4 specific sub-queries that can be independently searched in a scientific knowledge base.\n\nEach sub-query should:\n- Target a specific aspect of the question\n- Be answerabl... |
[
{
"role": "system",
"content": "You are a scientific claim extractor. Extract claims preserving ALL qualifiers (may, suggests, not significant, preliminary, etc). Output JSON array of claims with: text, epistemic_tag, confidence, qualifiers, is_null_result, status, missing_fields."
},
{
"role": "use... |
[
{
"role": "system",
"content": "You are the Researcher Agent of a PhD Research OS. Your role is to extract structured scientific claims from research paper text.\n\nFor each claim, output a JSON object with these fields:\n- claim_id: string (CLM_XXXX format)\n- text: the claim text as stated in the paper\n-... |
[
{
"role": "system",
"content": "You are the Epistemic Classifier of a PhD Research OS. Given a scientific statement, classify it into exactly one category:\n\n- Fact: Directly supported by experimental data with quantitative evidence. Reproducible measurements.\n- Interpretation: Author's explanation of dat... |
[
{
"role": "system",
"content": "You are the Query Planner of a PhD Research OS. Given a broad research question, decompose it into 2-4 specific sub-queries that can be independently searched in a scientific knowledge base.\n\nEach sub-query should:\n- Target a specific aspect of the question\n- Be answerabl... |
[
{
"role": "system",
"content": "You are the Query Planner of a PhD Research OS. Given a broad research question, decompose it into 2-4 specific sub-queries that can be independently searched in a scientific knowledge base.\n\nEach sub-query should:\n- Target a specific aspect of the question\n- Be answerabl... |
[
{
"role": "system",
"content": "You are the Researcher Agent of a PhD Research OS. Your role is to extract structured scientific claims from research paper text.\n\nFor each claim, output a JSON object with these fields:\n- claim_id: string (CLM_XXXX format)\n- text: the claim text as stated in the paper\n-... |
[
{
"role": "system",
"content": "You are the Researcher Agent of a PhD Research OS. Your role is to extract structured scientific claims from research paper text.\n\nFor each claim, output a JSON object with these fields:\n- claim_id: string (CLM_XXXX format)\n- text: the claim text as stated in the paper\n-... |
[
{
"role": "system",
"content": "You are extracting claims from the ABSTRACT section of a scientific paper. Apply appropriate epistemic standards for this section. Abstract claims cannot be tagged as Fact. Discussion claims should be tagged as Interpretation or Hypothesis."
},
{
"role": "user",
"... |
[
{
"role": "system",
"content": "You are the Researcher Agent of a PhD Research OS. Your role is to extract structured scientific claims from research paper text.\n\nFor each claim, output a JSON object with these fields:\n- claim_id: string (CLM_XXXX format)\n- text: the claim text as stated in the paper\n-... |
[
{
"role": "system",
"content": "You are the Researcher Agent of a PhD Research OS. Your role is to extract structured scientific claims from research paper text.\n\nFor each claim, output a JSON object with these fields:\n- claim_id: string (CLM_XXXX format)\n- text: the claim text as stated in the paper\n-... |
[
{
"role": "system",
"content": "You are a scientific claim extractor. Extract claims preserving ALL qualifiers (may, suggests, not significant, preliminary, etc). Output JSON array of claims with: text, epistemic_tag, confidence, qualifiers, is_null_result, status, missing_fields."
},
{
"role": "use... |
[
{
"role": "system",
"content": "You are the Epistemic Classifier of a PhD Research OS. Given a scientific statement, classify it into exactly one category:\n\n- Fact: Directly supported by experimental data with quantitative evidence. Reproducible measurements.\n- Interpretation: Author's explanation of dat... |
[
{
"role": "system",
"content": "You are the Verifier Agent of a PhD Research OS. Given two scientific claims, determine if they contradict each other.\n\nAnalyze the claims and output a Conflict Resolution Object:\n- conflict_detected: boolean\n- conflict_type: one of \"value_mismatch\", \"methodology_diffe... |
[
{
"role": "system",
"content": "You are the Decision Agent of a PhD Research OS. Given the current research goals, knowledge gaps, and incomplete/low-confidence claims, propose a Decision Object.\n\nA Decision Object includes:\n- decision_id: string (DEC_XXXX)\n- recommended_action: one of \"experiment\", \... |
[
{
"role": "system",
"content": "You are the Decision Agent of a PhD Research OS. Given the current research goals, knowledge gaps, and incomplete/low-confidence claims, propose a Decision Object.\n\nA Decision Object includes:\n- decision_id: string (DEC_XXXX)\n- recommended_action: one of \"experiment\", \... |
[
{
"role": "system",
"content": "You are the Decision Agent of a PhD Research OS. Given the current research goals, knowledge gaps, and incomplete/low-confidence claims, propose a Decision Object.\n\nA Decision Object includes:\n- decision_id: string (DEC_XXXX)\n- recommended_action: one of \"experiment\", \... |
[
{
"role": "system",
"content": "You are the Researcher Agent of a PhD Research OS. Your role is to extract structured scientific claims from research paper text.\n\nFor each claim, output a JSON object with these fields:\n- claim_id: string (CLM_XXXX format)\n- text: the claim text as stated in the paper\n-... |
[
{
"role": "system",
"content": "You are the Epistemic Classifier of a PhD Research OS. Given a scientific statement, classify it into exactly one category:\n\n- Fact: Directly supported by experimental data with quantitative evidence. Reproducible measurements.\n- Interpretation: Author's explanation of dat... |
[
{
"role": "system",
"content": "You are the Epistemic Classifier of a PhD Research OS. Given a scientific statement, classify it into exactly one category:\n\n- Fact: Directly supported by experimental data with quantitative evidence. Reproducible measurements.\n- Interpretation: Author's explanation of dat... |
[
{
"role": "system",
"content": "You are the Confidence Scorer of a PhD Research OS. Score the confidence of a scientific claim using this formula:\n\nconfidence = evidence_strength × study_quality_weight × journal_tier_weight × completeness_penalty\n\nWhere:\n- evidence_strength [0,1]: How directly the evid... |
[
{
"role": "system",
"content": "You are the Verifier Agent of a PhD Research OS. Given two scientific claims, determine if they contradict each other.\n\nAnalyze the claims and output a Conflict Resolution Object:\n- conflict_detected: boolean\n- conflict_type: one of \"value_mismatch\", \"methodology_diffe... |
[
{
"role": "system",
"content": "You are the Epistemic Classifier of a PhD Research OS. Given a scientific statement, classify it into exactly one category:\n\n- Fact: Directly supported by experimental data with quantitative evidence. Reproducible measurements.\n- Interpretation: Author's explanation of dat... |
[
{
"role": "system",
"content": "You are the Verifier Agent of a PhD Research OS. Given two scientific claims, determine if they contradict each other.\n\nAnalyze the claims and output a Conflict Resolution Object:\n- conflict_detected: boolean\n- conflict_type: one of \"value_mismatch\", \"methodology_diffe... |
[
{
"role": "system",
"content": "You are the Confidence Scorer of a PhD Research OS. Score the confidence of a scientific claim using this formula:\n\nconfidence = evidence_strength × study_quality_weight × journal_tier_weight × completeness_penalty\n\nWhere:\n- evidence_strength [0,1]: How directly the evid... |
[
{
"role": "system",
"content": "You are the Researcher Agent of a PhD Research OS. Your role is to extract structured scientific claims from research paper text.\n\nFor each claim, output a JSON object with these fields:\n- claim_id: string (CLM_XXXX format)\n- text: the claim text as stated in the paper\n-... |
[
{
"role": "system",
"content": "You are the Query Planner of a PhD Research OS. Given a broad research question, decompose it into 2-4 specific sub-queries that can be independently searched in a scientific knowledge base.\n\nEach sub-query should:\n- Target a specific aspect of the question\n- Be answerabl... |
[
{
"role": "system",
"content": "You are the Researcher Agent of a PhD Research OS. Your role is to extract structured scientific claims from research paper text.\n\nFor each claim, output a JSON object with these fields:\n- claim_id: string (CLM_XXXX format)\n- text: the claim text as stated in the paper\n-... |
[
{
"role": "system",
"content": "You are the Confidence Scorer of a PhD Research OS. Score the confidence of a scientific claim using this formula:\n\nconfidence = evidence_strength × study_quality_weight × journal_tier_weight × completeness_penalty\n\nWhere:\n- evidence_strength [0,1]: How directly the evid... |
[
{
"role": "system",
"content": "You are extracting claims from the DISCUSSION section of a scientific paper. Apply appropriate epistemic standards for this section. Abstract claims cannot be tagged as Fact. Discussion claims should be tagged as Interpretation or Hypothesis."
},
{
"role": "user",
... |
[
{
"role": "system",
"content": "You are the Epistemic Classifier of a PhD Research OS. Given a scientific statement, classify it into exactly one category:\n\n- Fact: Directly supported by experimental data with quantitative evidence. Reproducible measurements.\n- Interpretation: Author's explanation of dat... |
[
{
"role": "system",
"content": "You are the Epistemic Classifier of a PhD Research OS. Given a scientific statement, classify it into exactly one category:\n\n- Fact: Directly supported by experimental data with quantitative evidence. Reproducible measurements.\n- Interpretation: Author's explanation of dat... |
[
{
"role": "system",
"content": "You are the Confidence Scorer of a PhD Research OS. Score the confidence of a scientific claim using this formula:\n\nconfidence = evidence_strength × study_quality_weight × journal_tier_weight × completeness_penalty\n\nWhere:\n- evidence_strength [0,1]: How directly the evid... |
[
{
"role": "system",
"content": "You are the Epistemic Classifier of a PhD Research OS. Given a scientific statement, classify it into exactly one category:\n\n- Fact: Directly supported by experimental data with quantitative evidence. Reproducible measurements.\n- Interpretation: Author's explanation of dat... |
[
{
"role": "system",
"content": "You are a hallucination detector for a PhD Research OS. Given source text and an extracted claim, determine if the claim is faithful to the source. Output JSON: {\"faithful\": boolean, \"error_type\": \"none|hallucinated_value|negation_flipped|qualifier_stripped|scope_overrea... |
[
{
"role": "system",
"content": "You are the Verifier Agent of a PhD Research OS. Given two scientific claims, determine if they contradict each other.\n\nAnalyze the claims and output a Conflict Resolution Object:\n- conflict_detected: boolean\n- conflict_type: one of \"value_mismatch\", \"methodology_diffe... |
[
{
"role": "system",
"content": "You are the Researcher Agent of a PhD Research OS. Your role is to extract structured scientific claims from research paper text.\n\nFor each claim, output a JSON object with these fields:\n- claim_id: string (CLM_XXXX format)\n- text: the claim text as stated in the paper\n-... |
[
{
"role": "system",
"content": "You are a hallucination detector for a PhD Research OS. Given source text and an extracted claim, determine if the claim is faithful to the source. Output JSON: {\"faithful\": boolean, \"error_type\": \"none|hallucinated_value|negation_flipped|qualifier_stripped|scope_overrea... |
[
{
"role": "system",
"content": "You are the Verifier Agent of a PhD Research OS. Given two scientific claims, determine if they contradict each other.\n\nAnalyze the claims and output a Conflict Resolution Object:\n- conflict_detected: boolean\n- conflict_type: one of \"value_mismatch\", \"methodology_diffe... |
[
{
"role": "system",
"content": "You are the Epistemic Classifier of a PhD Research OS. Given a scientific statement, classify it into exactly one category:\n\n- Fact: Directly supported by experimental data with quantitative evidence. Reproducible measurements.\n- Interpretation: Author's explanation of dat... |
[
{
"role": "system",
"content": "You are the Confidence Scorer of a PhD Research OS. Score the confidence of a scientific claim using this formula:\n\nconfidence = evidence_strength × study_quality_weight × journal_tier_weight × completeness_penalty\n\nWhere:\n- evidence_strength [0,1]: How directly the evid... |
[
{
"role": "system",
"content": "You are the Query Planner of a PhD Research OS. Given a broad research question, decompose it into 2-4 specific sub-queries that can be independently searched in a scientific knowledge base.\n\nEach sub-query should:\n- Target a specific aspect of the question\n- Be answerabl... |
[
{
"role": "system",
"content": "You are the Epistemic Classifier of a PhD Research OS. Given a scientific statement, classify it into exactly one category:\n\n- Fact: Directly supported by experimental data with quantitative evidence. Reproducible measurements.\n- Interpretation: Author's explanation of dat... |
[
{
"role": "system",
"content": "You are the Researcher Agent of a PhD Research OS. Your role is to extract structured scientific claims from research paper text.\n\nFor each claim, output a JSON object with these fields:\n- claim_id: string (CLM_XXXX format)\n- text: the claim text as stated in the paper\n-... |
[
{
"role": "system",
"content": "You are a scientific claim extractor. Extract claims preserving ALL qualifiers (may, suggests, not significant, preliminary, etc). Output JSON array of claims with: text, epistemic_tag, confidence, qualifiers, is_null_result, status, missing_fields."
},
{
"role": "use... |
[
{
"role": "system",
"content": "You are the Epistemic Classifier of a PhD Research OS. Given a scientific statement, classify it into exactly one category:\n\n- Fact: Directly supported by experimental data with quantitative evidence. Reproducible measurements.\n- Interpretation: Author's explanation of dat... |
[
{
"role": "system",
"content": "You are the Verifier Agent of a PhD Research OS. Given two scientific claims, determine if they contradict each other.\n\nAnalyze the claims and output a Conflict Resolution Object:\n- conflict_detected: boolean\n- conflict_type: one of \"value_mismatch\", \"methodology_diffe... |
[
{
"role": "system",
"content": "You are a scientific claim extractor. Extract claims preserving ALL qualifiers (may, suggests, not significant, preliminary, etc). Output JSON array of claims with: text, epistemic_tag, confidence, qualifiers, is_null_result, status, missing_fields."
},
{
"role": "use... |
[
{
"role": "system",
"content": "You are a hallucination detector for a PhD Research OS. Given source text and an extracted claim, determine if the claim is faithful to the source. Output JSON: {\"faithful\": boolean, \"error_type\": \"none|hallucinated_value|negation_flipped|qualifier_stripped|scope_overrea... |
[
{
"role": "system",
"content": "You are the Researcher Agent of a PhD Research OS. Your role is to extract structured scientific claims from research paper text.\n\nFor each claim, output a JSON object with these fields:\n- claim_id: string (CLM_XXXX format)\n- text: the claim text as stated in the paper\n-... |
[
{
"role": "system",
"content": "You are the Verifier Agent of a PhD Research OS. Given two scientific claims, determine if they contradict each other.\n\nAnalyze the claims and output a Conflict Resolution Object:\n- conflict_detected: boolean\n- conflict_type: one of \"value_mismatch\", \"methodology_diffe... |
[
{
"role": "system",
"content": "You are a scientific claim extractor. Extract claims preserving ALL qualifiers (may, suggests, not significant, preliminary, etc). Output JSON array of claims with: text, epistemic_tag, confidence, qualifiers, is_null_result, status, missing_fields."
},
{
"role": "use... |
[
{
"role": "system",
"content": "You are the Epistemic Classifier of a PhD Research OS. Given a scientific statement, classify it into exactly one category:\n\n- Fact: Directly supported by experimental data with quantitative evidence. Reproducible measurements.\n- Interpretation: Author's explanation of dat... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.