Claim veracity and fact-checking
MIS-01Verified claims with ratings, evidence trails, and independent methodology.
truthfulnessalignment
Misinformation Datasets — Cognitive & Semantic Structures
These datasets provide the metric rods the machine uses to evaluate misinformation. Each rod is a small, structured JSON file containing authoritative examples that help detect:
- ›semantic drift
- ›chromatic bias
- ›temporal distortion
- ›narrative scope collapse
- ›synthetic media manipulation
- ›source credibility issues
- ›amplification dynamics
- ›fact-checking stability
Metric rods are tiny, precise, and machine-readable — perfect for alignment challenges.
Dataset family
Semantic drift and terminological hijacking
MIS-02Terms whose meaning shifts under political or commercial pressure.
cohesiontruthfulness
Narrative scope collapse
MIS-03Multi-causal situations compressed into a single salient element.
cohesionquantification
Temporal distortion and out-of-date framing
MIS-04Outdated or re-dated content; zombie claims reactivated.
truthfulnessquantification
Source credibility and citation chains
MIS-05Provenance, expertise, independence, transparency.
truthfulnessalignment
Deepfakes and synthetic media
MIS-06Manipulated or generated media; forensic signals and provenance markers.
truthfulnessalignment
Amplification dynamics on platforms
MIS-07Cascade shapes, virality, recommender behavior.
quantificationcohesion
Chromatic bias in emotive framing
MIS-08Sentiment, priming, outrage load, tone.
cohesionquantification
Rod-set downloads
Each category resolves to a metric rod set — a collection of JSON rods the engine reads directly.
- Fact-checking & veracity rods↓ JSON
- Semantic drift & framing rods↓ JSON
- Temporal distortion rods↓ JSON
- Source credibility rods↓ JSON
- Amplification & synthetic media rods↓ JSON
These rod sets calibrate the machine's alignment logic for the Misinformation Challenge.