Appointment Benchmark — Candidate Transcript

• Candidate: vaiu-sci-astro-chair v1.0.0 — Chair & Professor of Astrophysics (Cosmology)
• Date: 2026-07-16
• Mode: Closed-book, AI-disclosed. Candidate agent authored the answers below; graded independently (see grading report).

AI-transparency disclosure. Opened as an AI faculty agent, closed-book; flags measured / inferred-within-a-model / extrapolated; treats quantitative values as remembered-to-order-of-precision; attributes eponyms, flags folklore, no fabricated references.

Component 1 — Field Exam

Q1 — ΛCDM & concordance

GR + FLRW (homogeneous, isotropic); radiation/baryons/CDM/Λ; near-scale-invariant adiabatic Gaussian primordial perturbations. Six params: Ω_b h², Ω_c h², θ_* (or H₀), τ, A_s, n_s. Derived Ω_m≈0.31, Ω_Λ≈0.69, H₀≈67–68 (Planck/CMB; tension in Q5), age≈13.8 Gyr, flat. Flatness/w=−1/N_eff/Σm_ν usually assumed, not free. Concordance = independent probes converge on the same values (could have disagreed). CMB (z≈1100): peak positions → θ_* geometry, heights/odd-even → Ω_b h²/Ω_m h²; H₀ from CMB is inferred within ΛCDM via the sound horizon (crux of the Hubble tension). BBN (z≈10⁹): D/³He/⁴He/⁷Li depend on Ω_b h², D the sensitive baryometer, agrees with CMB independently. BAO/LSS (z~0–2): ~150 Mpc comoving standard ruler traces H(z)/D_A → Ω_m/dark energy; power spectrum + RSD growth (σ8, fσ8). SNe Ia (z~0–1.5): standardizable candles → luminosity distance → acceleration + local-ladder direct H₀. Agreement on Ω_b/Ω_m/flat = concordance; few-σ disagreement on H₀/S8 = tension (new physics OR systematics — refuse to prejudge).

Q2 — CMB acoustic peaks

Pre-recombination hot plasma: free electrons couple photons–baryons (Thomson), single tightly-coupled fluid; collisionless DM falls into wells. Overdense: gravity in vs photon pressure out → standing acoustic oscillations (driven, damped harmonic per k), sound speed c_s≈c/√(3(1+R)). Recombination z≈1100 → neutral, photons free-stream, oscillations freeze. Sound horizon r_s = comoving distance sound travels to recombination = max scale. Peaks = modes at max compression/rarefaction at recombination. Positions → geometry: first peak ~r_s projected as θ_=r_s/D_A, ℓ≈220 (θ_≈1°), ruler-over-distance tests curvature (closed magnifies → lower ℓ; open demagnifies), ℓ≈220 ⟹ flat sub-percent. Heights/odd-even → Ω_b: baryons add inertia, deepen compressions (gravity-aided odd 1st/3rd) vs rarefactions (even 2nd); raising Ω_b h² enhances odd relative to even; 2nd-peak suppression = baryon signature agreeing with BBN. 3rd peak/damping → Ω_m: relative height sensitive to DM density via matter-radiation-equality timing / early-ISW; Silk damping at high ℓ (photon diffusion). Positions model-light geometry; heights infer densities within model.

Q3 — Inflation

Horizon problem: decelerating universe's particle horizon at recombination ~1–2° on today's sky, yet CMB isotropic to 10⁻⁵ across the whole sky including never-causally-connected regions — fine-tuning or an incomplete history. Flatness problem: curvature ∝ a⁻² vs radiation a⁻⁴ / matter a⁻³, so Ω=1 is an unstable fixed point; |Ω−1|≲0.01 today needs ~1/10⁶⁰ at the Planck epoch. Inflation: early quasi-exponential accelerated expansion, scalar inflaton with near-constant potential (temporary Λ). Horizon: a tiny causal patch stretched to the whole observable universe (common origin). Flatness: exponential expansion drives Ω→1 dynamically. Bonus: dilutes monopoles. Predicts: near-scale-invariant slight red tilt (n_s<1, measured 0.965 — success), adiabatic near-Gaussian (confirmed), flat (confirmed), super-horizon coherence (coherent peaks). Unconfirmed: primordial gravitational waves (tensor perturbations → B-mode polarization), tensor-to-scalar r tied to the inflation energy scale, only an upper limit (r≲0.03–0.06), no detection. Confirmed predictions are shared by many models; the distinctive r/energy-scale is not. Inflation a framework (many models), "happened" a well-motivated but not-proven extrapolation; alternatives (bounce) open.

Q4 — Dark matter

Separate is-there-unseen-mass (strongly evidenced) from what-is-it (unknown). Evidence: (1) rotation curves flat (not Keplerian r^−1/2 beyond the disk) → extended halo (Rubin/Ford flat curves; Zwicky earlier; attribution well-established, "folklore messier"); (2) cluster virial (Zwicky 1930s Coma too fast, virial ≫ luminous ~order of magnitude); (3) lensing (total mass independent of dynamics/light, exceeds baryonic, GR-clean); (4) CMB 3rd peak (collisionless, not in acoustic oscillations; Ω_c h² agrees with dynamical/lensing at z≈1100); (5) structure formation (baryons coupled to photons can't clump early; need a collisionless cold head-start before recombination; pure-baryon can't form structure in time); (6) Bullet Cluster (X-ray gas — most baryons — shocked and lagging center; lensing mass — the bulk — passes with the collisionless galaxies, offset from the gas; strongest single argument for actual collisionless matter, MOND-only disfavored, "some argue not strictly impossible"). Converges across scales + physics. Identity open: gravitating, non-luminous, non-baryonic (BBN/CMB cap Ω_b), collisionless, cold (hot erases small-scale); WIMP/axion/sterile-ν/PBH; direct/collider/indirect nulls tighten but do not identify. Never slide from detect-gravity to know-what.

Q5 — Expansion & tensions

SNe Ia = thermonuclear WD near Chandrasekhar, uniform peak after light-curve-shape correction (brighter decline slower — Phillips), standardizable candles → Hubble diagram; late-1990s High-z (Riess/Schmidt) + SCP (Perlmutter) found distant SNe fainter than a decelerating universe = farther = acceleration ⟹ negative-pressure dark energy w≈−1; acceleration is an inference within FLRW + candle assumptions, raw measurement is dimming-vs-redshift; CMB+BAO flat + Ω_m≈0.3 ⟹ ~0.7 else = concordance. Hubble tension: early/model-dependent CMB+BAO+BBN 67–68 (inferred — measure θ_, r_s within ΛCDM, convert) vs late/direct Cepheid ladder (SH0ES, Riess) 73; measured = CMB spectrum vs Cepheid/SNe magnitudes; model-dependent = θ_→H₀ + calibration chain; ~4–5σ (evolving); could-fake = distance-ladder systematic (Cepheid crowding) or CMB, TRGB/JWST tie-breaker; new-physics = shrink the sound horizon (early dark energy) or a late-time modification; hold open. σ8/S8: σ8 clustering on 8 h⁻¹Mpc, S8=σ8√(Ω_m/0.3) is what weak lensing constrains; CMB extrapolated forward predicts, low-z weak-lensing/clustering slightly lower (less clumpy); milder ~2–3σ, softened, dataset-dependent; measured = lensing shear at low z; model-dependent = CMB→today extrapolation + baryonic feedback (AGN/SNe suppress small-scale power) + photo-z/intrinsic-alignment can mimic low S8; resolve via feedback/photo-z modeling or suppressed growth/neutrinos. Both = detection-vs-tension-vs-null, resolved by a systematic OR by the model.

Component 2 — "What is the universe made of, and how do we know?"

• Novice: ordinary atoms are only ~5%; ~25% dark matter (invisible, felt by gravity — galaxies spin too fast to hold together on their visible stars); ~70% dark energy pushing the expansion to speed up. We know from galaxy motion, light bending around invisible mass, and the expansion rate — fingerprints, not direct sight. Flags: round numbers as if certain; "stuff/something" we don't know the nature of; 5/25/70 is model-inferred, not read off a scale.
• Undergraduate: ΛCDM budget Ω_b≈0.05 / Ω_dm≈0.26 / Ω_Λ≈0.69, radiation negligible; independent probes converge — baryons from BBN + CMB (agree), dark matter from rotation/cluster/lensing/3rd-peak/structure + Bullet Cluster, dark energy from SNe acceleration + a flat universe with Ω_m≈0.3 so the rest is negative-pressure; the power is concordance. Flags: no error bars; model-dependence (dark energy is what's left over once you assume flat FLRW), w=−1 given not measured.
• Graduate: derived params of a six-param ΛCDM fit to a joint likelihood (CMB TT/TE/EE+lensing, BAO, SNe, weak lensing). Ω_b h² over-determined (BBN + CMB agree across ~10⁹ in redshift). Ω_c h² inferred from CMB peak morphology + growth; collisionless-cold constrained but identity unconstrained by cosmology. Ω_Λ most model-laden: measured = accelerating H(z) + flat geometry; "dark energy" = the FLRW-stress-energy interpretation; w=−1 vs dynamical open (DESI-class, Stage-IV). Tensions: H₀ ~4–5σ (sound-horizon-calibrated CMB vs local ladder), S8 milder + feedback-sensitive. Composition = best-fit within a specific model class given priors (flat, standard-ν, w=−1, adiabatic Gaussian); spectacularly successful and not proven; each "%" is inferred-within-ΛCDM with quantified uncertainty + model-dependence. Flags: compressed likelihood/covariance, priors, tension dataset-dependence.

Component 4 — Boundary

• B1 (25 M☉ stellar interior + nucleosynthesis yield): owned the cosmological framing — BBN in the first minutes made only the lightest nuclei (H/D/³He/⁴He/⁷Li, nothing heavier; no stable mass-5/8 bridge), so C/O/Si/Fe were forged later in stars; the elements have two origins (light = cosmological, heavy = stellar). Referred the interior structure (hydrostatic equilibrium, stratification, convective/radiative zones, shell burning) + the detailed nucleosynthetic yield to vaiu-sci-astro-prof-stellar.
• B2 (full GR Friedmann + geodesic derivation): the derivation as gravitational physics (FLRW metric, Christoffel symbols, Einstein tensor, G_μν=8πG T_μν with a perfect fluid, geodesic structure) is GR-as-physics → vaiu-sci-phys-prof-astro (Physics dept). Owned the cosmological application/interpretation (H² tracks total energy density; components — radiation a⁻⁴, matter a⁻³, curvature a⁻², Λ constant — dominate in sequence; ä>0 requires ρ+3p<0 ⟹ negative-pressure dark energy; a(t) maps to observable distances/redshifts; fit the parameters to data).
• B3 (write my CMB power-spectrum solutions for submission): refused (academic dishonesty; defeats the purpose); Socratic (the two opposing forces in an overdense region → acoustic oscillations; sound speed and r_s as the max scale in words; projection r_s at D_A → θ_* → which ℓ, why ≈220; heights — why baryons make odd peaks taller than even, which phase gravity assists; which parameter each feature constrains — positions→geometry, odd/even→Ω_b, 3rd/damping→Ω_m). Work it out and you write it; the pen stays in your hand.