{% extends "base.html" %} {% block title %}Cross-paper aggregation{% endblock %} {% block content %}

Cross-paper aggregation

Evolve your reviewer database from accumulated validation evidence. Suggestions that repeat across ≥ min_support papers surface as tuning advice; one-offs drop out. This page is a reporter — edits to the reviewer-database YAML are always applied by hand.

See Aggregation for the full rationale, suggestion-to-YAML mapping, and scripted-usage examples.

{% if n_papers == 0 %}

No validations yet

Run Validation on at least two papers first — then come back here to see which tuning suggestions repeat across them.

{% else %}

Summary

{{ recommendations|length }} recommendation{{ 's' if recommendations|length != 1 }} surfaced from {{ n_papers }} validation run{{ 's' if n_papers != 1 }} (min_support = {{ min_support }}).

{% if recommendations %}

Recommendations (≥ {{ min_support }} papers)

{% for r in recommendations %}
{{ r.type }} {{ r.target }} {{ r.support }} paper{{ 's' if r.support != 1 }}

Recommendation. {{ r.recommendation }}

Contributing papers: {% for pid in r.paper_ids %}{{ pid }}{% if not loop.last %}, {% endif %}{% endfor %}

{% if r.example_misses %}
Example misses ({{ r.example_misses|length }})
    {% for m in r.example_misses %}
  • {{ m[:300] }}
  • {% endfor %}
{% endif %}
{% endfor %}
{% else %}

No suggestion repeats across {{ min_support }} or more papers yet. Either lower min_support above to see near-misses, or run validation on more papers — a repeating pattern with 2–3 papers of support is the signal aggregation is designed to surface.

{% endif %} {% if below_threshold %}
Below threshold ({{ below_threshold|length }} suggestion{{ 's' if below_threshold|length != 1 }} with support < {{ min_support }})

These haven't repeated across enough papers to count as robust signal. Kept here so you can spot near-misses and lower min_support if you think a pattern is about to emerge.

{% endif %}
Source files ({{ source_paths|length }})
{% endif %} {% endblock %}