312 questions and 903 question–evidence pairs
Research question
Can a decision model identify whether evidence answers an open question?
Approach
A model checks question–evidence pairs using a fixed rubric. Independent reviews and later rule changes are considered separately.
What is available so far
The documented first run of rubric 4.4 provides proposed answer states and review reasons. The subsequent 4.5 run was rejected because of an overly restrictive rule.
Limits of the findings
Proposals are not confirmed answers. Uncertainty, addressee matching and rule changes limit the findings.
Next step
Evaluate rules against counter-reviews before adopting new proposed answer states.
Basis of this account
- First Raster g 4.4 run dated 1 October 2026
- Results and rejection of the subsequent 4.5 run
This page summarises the documented state of the work. A full standalone presentation of the results has not yet been prepared.
The data snapshot refers to the underlying investigation, not the date of this overview.