Skip to content

Latest commit

 

History

History
43 lines (32 loc) · 2.63 KB

File metadata and controls

43 lines (32 loc) · 2.63 KB

Research contract

This contract applies to every candidate, experiment, report, and automated component.

Evidence

  1. Every consequential factual claim must identify its source and precise source location.
  2. AI-generated summaries are unverified until checked against the source.
  3. Secondary sources may orient retrieval but do not replace primary evidence for scientific claims.
  4. Missing, inaccessible, retracted, corrected, or disputed evidence must remain visible.
  5. The corpus must have documented inclusion, exclusion, coverage, and stopping rules.

Reasoning

  1. Records must distinguish observation, statistical relationship, model inference, causal hypothesis, and speculation.
  2. Every promoted question must include competing explanations and evidence against novelty.
  3. Terminology similarity is a lead, not proof that phenomena or mechanisms are equivalent.
  4. Importance, novelty, tractability, and available-data quality are separate judgments.
  5. Uncertainty and reviewer disagreement must not be collapsed into a single confidence score.

Experiments

  1. The question, hypotheses, data rules, primary outcomes, holdouts, tests, and falsification criteria must be preregistered before final evaluation.
  2. Training, exploration, validation, and final holdout data must be separated where the design permits.
  3. Baseline reproduction precedes novel analysis.
  4. Leakage, confounding, robustness, measurement error, and alternative explanations must be tested.
  5. Deviations from preregistration must be timestamped and labeled exploratory.
  6. Negative results and failed experiments remain in the record.
  7. Reproducibility includes code, configuration, environment, data provenance, and permitted acquisition steps.

Claims and review

  1. Correlation cannot be described as causation.
  2. “Novel” means no prior work was found after a documented adversarial search; it is never an absolute guarantee.
  3. Computational evidence cannot be presented as physical or clinical validation.
  4. Expert scrutiny is required when interpretation depends materially on domain knowledge unavailable within the evidence record.
  5. Publication, public claims, human-subject work, restricted data, laboratory work, or consequential external action requires a separate explicit decision.

Automation boundary

  1. Automation may retrieve, structure, compare, generate candidates, and run declared computations.
  2. Automation may not silently expand scope, suppress contrary evidence, select a final question, or promote a scientific claim.
  3. The system must preserve an auditable path from source to claim to candidate to experiment to conclusion.