---------------------------
Exoplanet Validation Worlds
---------------------------

This directory contains a single Notation3 file that encodes a case 
study about validating exoplanet candidates from a transit survey 
pipeline in four different “worlds” of reasoning. Each candidate is 
described by a prior occurrence rate for that type of planet, and by 
the pipeline’s sensitivity and specificity. The reference world W0 
applies Bayes’ rule to combine these into a posterior probability 
that a detected transit-like signal is a real planet, and confirms the 
candidate only if that probability is high enough. W1 behaves like a 
naive view that looks mostly at sensitivity and treats the pipeline as 
“great” whenever it rarely misses planets, without fully accounting 
for occurrence or false positives. W2 uses a simple heuristic gate on 
occurrence, sensitivity and specificity. W3 is a cautious Bayes world 
that uses the same posterior as W0 but demands a stricter confirmation 
threshold.

From these inputs, the core rules compute, for each candidate and 
world, whether the signal is confirmed or rejected. An ARC layer then 
derives an Answer node summarising how the worlds agree and disagree on 
four illustrative candidates (from a very rare long-period regime to a 
common hot Neptune with reduced specificity), a Reason node explaining 
how the worlds relate to Bayes’ rule, and five Check nodes that 
capture relationships like “only the naive world confirms a very rare 
candidate” and “the cautious Bayes world is at least as 
conservative as the reference Bayes world on these examples”. All ARC 
triples are derived from the same numeric core, so changing the 
candidates or world thresholds will update the Answer, Reason and 
Checks automatically.
