Role
Benchmark
Sets the main human baseline before the wider launch-phase and adversarial comparisons.
Slices
Halves / thirds
Splits the same corpus multiple ways so the result has to survive internal resampling.
Question
Operating band
Asks where the measured floor and ceiling sit under first-draft conditions.
Evidence Frame
What it is
The Solo Dataset is the core human benchmark for PRM: first-draft, unaided, non-repeating original work measured as aggregate corpus behavior.
The public page describes structure, measurement layers, and aggregate results. It does not publish protected source text, source-level mappings, or a downloadable corpus.
Structure
- Full catalog: the complete solo writing body used for internal analysis.
- PRM Half 1: the earlier large-scale baseline.
- PRM Half 2 / The Sequel: the later half-corpus segment.
- PRM Third 1: early-third behavior.
- PRM Third 2: middle-third behavior.
- PRM Third 3 / Launch Phase: final-third behavior inside The Sequel.
Those cuts are broad enough to make cherry-picking harder. The public evidence is about how large blocks behave, not how a private excerpt performs.
Why it matters
The Solo Dataset is where PRM’s internal evolution can be tested.
Instead of comparing one highlight against another highlight, the analysis compares large chronological blocks. That matters because the central findings are not framed as isolated peaks. The important question is whether later writing raises the operating band across scale, slice size, and metric family.
What the comparison tests
The public Solo view focuses on aggregate contrasts between earlier and later PRM blocks. Half-corpus and third-corpus splits let the analysis ask three related questions:
- Does the later half separate from the earlier half?
- Does the final third explain the acceleration inside the later half?
- Does the later floor rise above the earlier ceiling?
Those questions connect the Solo Dataset to The Sequel, Launch Phase, and New Floor Above Old Ceiling findings.
Major findings
- PRM Half 1 establishes the earlier operating band.
- PRM Half 2 / The Sequel separates at half-corpus scale.
- PRM Third 3 / Launch Phase explains the late acceleration inside The Sequel.
- New Floor Above Old Ceiling frames the raised-baseline interpretation.
How to read the charts
Start with the Sequel gap chart. It shows whether the later half keeps its advantage across measurement windows, separating stable corpus behavior from a narrow spike.
Then read the metric-delta chart as a family-by-family comparison between The Sequel and PRM Half 1. Floor / center / peak shows whether the later block is improving only at the top end or lifting the whole distribution. Stability asks whether strength survives across slice sizes, and the heatmap shows where the strongest deltas appear.
Public-safe limits
The public Solo page does not expose raw writing, protected source excerpts, private file manifests, third-party source labels, artist names, album titles, or song titles.
The page explains what was measured and how the public aggregate charts should be interpreted. It does not publish or package a public corpus.
Public-safe boundary
Public pages show aggregate evidence, metric behavior, method provenance, and corpus structure. Protected text, identities, source titles, and reconstructable mappings stay private.