C4 — Has selection quality improved across cohorts?

The 10.0 pipeline in brief. ~2,200 people applied. Each applicant went through three stages:

1. Stage 1 — submitted background / experience / motivation, picked which research tracks they were interested in (Empirical, Policy & Strategy, Technical Governance, Theory, Compute Infrastructure). An LLM screen filtered out applicants who clearly didn't meet a minimum bar, and produced advisory per-stream recommendations.
2. Stage 2 — applicants who passed Stage 1 had their materials scored by LLM-graded rubrics. The empirical track used a composite score combining Research Skills, Technical Execution (split into MLE, SWE, Math sub-scores), and Soft Skills. The top ~600 by composite advanced to Stage 3.
3. Stage 3 — applicants chose specific mentors / "streams" to apply to. Each stream reviewed its applicants and produced a ranked list. Top-ranked applicants got offers; lower-ranked got waitlisted. ~120 offers were made.

For the empirical track, the composite formula is 0.50·RS + 0.35·TE + 0.15·SS, where TE = 0.50·MLE + 0.30·SWE + 0.20·Math. A "relevance multiplier" (Direct=1.0 / Adjacent=0.85 / Distant=0.60) is applied to Research Skills based on how the applicant's experience matches the streams they applied to.

Outcome definitions used throughout these analyses:

• is_ranked (primary outcome) — applicant was ranked by ≥1 stream. This is the cleanest signal of "the selection process picked this person." Not the same as "received an offer" — offer count is bounded by cohort size (~120), but rank count reflects quality independently of capacity.
• is_invited_to_worktest (secondary outcome) — applicant was engaged by ≥1 stream in any way: invited to a work test, invited to an interview, ranked, or sent the Megastream takehome. Strict superset of is_ranked. One level above is_ranked in the funnel.
• passed_mentors_bar — applicant was offered or waitlisted. In 10.0, this equals is_ranked exactly (every ranked person got either an offer or a waitlist slot).

• Different mentor-eval instruments. 6.0 is end-of-program; 7.0–9.0 are mid-program. The 9.0 instrument is richer (added reference-class rank, hire/pub likelihood). We use the four-dimension composite that's common across cohorts.
• SRP/FRP rubric changes. 7.0: ToC + Activities. 8.0: ToC + Progress + Activities. 9.0/FRP: research success, AI risk reduction, researcher ability, symposium. Raw scores are NOT comparable; we report within-cohort percentiles where needed.
• Cohort size scales. 7.0 had ~75 fellows; 10.0 has ~120. More fellows could dilute average quality (regression to the mean) — or not, if applicant pool grew proportionally.
• Publication recency. 6.0 alumni have had ~2 years to publish; 9.0 alumni have had ~6 months. Cross-cohort publication-rate comparisons partly reflect time-to-publish, not just program quality.
• 10.0 isn't here yet. The program just started; we have no in-program data on 10.0 fellows.