Calibration
Management guidance graded against the filed number. Each row is one executive; the grade accrues to the company's management until the attribution is upgraded. A calibration profile is a SOURCE-accuracy ledger. It confers no tier, no weight, no conviction, no corroboration.
| company | executive | graded (n) | HIT | MISS | mean bias | regime note |
|---|---|---|---|---|---|---|
| HOUSE | House thesis — temporal claim the retired temporal sentence (moves down over time — an open question, not a house claim) · pre-registered test H1 | 2 | 0 | 2 | — | first test NOT SUPPORTED, 2026-09-07 — falsifier fired on order, not on span; a second test is owed |
| GEV | Michael Lapides Vice President of Investor Relations, GE Vernova | 0 | 0 | 0 | — | none yet — n too small for a note |
| GEV | Ken Parks CFO, GE Vernova | 2 (30 pending) | 0 | 2 | +11.1% | 2 grades, one quarter (Q1-2026), one regime (gas-turbine demand rising, Wind contracting): … |
| GEV | Scott Strazik CEO, GE Vernova | 0 (3 pending) | 0 | 0 | — | none yet — n too small for a note |
| MU | Satya Kumar Corporate Vice President, Investor Relations and Treasury | 0 | 0 | 0 | — | none yet — n too small for a note |
| MU | Sanjay Mehrotra Chairman, President and Chief Executive Officer | 0 (7 pending) | 0 | 0 | — | none yet — n too small for a note |
| MU | Mark Murphy Executive Vice President and Chief Financial Officer | 16 (12 pending) | 2 | 14 | +14.8% | 16 grades within one regime (rising HBM demand); +14.8% mean bias is … |
| NVDA | Toshiya Hari Vice President-Investor Relations & Strategic Finance, NVIDIA Corp. | 0 | 0 | 0 | — | none yet — n too small for a note |
| NVDA | Jen Hsun Huang Founder, President, Chief Executive Officer & Director, NVIDIA Corp. | 0 | 0 | 0 | — | none yet — n too small for a note |
| NVDA | Colette M. Kress Chief Financial Officer & Executive Vice President, NVIDIA Corp. | 14 (15 pending) | 10 | 4 | +1.7% | Regime test UNTESTABLE for FY2024: every graded quarter is an up-guide (guide … |
mean bias: mean of (actual − point) / (high − low) over graded RANGE claims; computed separately, never graded (R-MGMT-VOICE-2). Read from voices/management/<TICKER>/<person>.json at build.
The mentor's priors, graded
Before each first filing read, a mentor names the signature the filing should show. The read then rewrites the target. Running record: 3 for 8.
- hits — GEV prepayments, TSEM capacity_reservation, VRT lead_time
- misses — AMAT lead_time, CEG take_or_pay, MRVL capacity_reservation (inverted), AAOI prepayments (falsified), ASML lead_time (not supported)
- unscored — TSM and INTC capacity_reservation (signature yes, rung no) (right signature, wrong layer)
Read from scout/rulings.json. A prior is a hypothesis, never evidence; the record grows one line per first read and is never averaged.
Two catches from the last review
The most recent scoring rep was drafted unattended and reviewed before admission. Two corrections were made at the review, and both are recorded on the ledger rather than smoothed over:
- A score was held, not raised. The draft read a peer's filing as support; the review read the same filing as adverse for the name (the constraint is real, the lock-in is absent) and held the thesis factor where it was. (review of 2026-09-06)
- A catalyst was scored on the canon scale, not on enthusiasm. The draft gave a confirmed near-dated event the top of the range; the scale caps proximity, and the review applied the cap. A rounding rule was written the same day: R-ROUND-TRUNCATE: quality and conviction are INTEGERS; a fractional binding factor TRUNCATES (3.75 → 3); the fraction is recorded beside the integer; a raise requires the BINDING FACTOR ITSELF to move, never the rounding.
Pre-registered: the 2026-09-30 print
Written 2026-09-06, before the event. THE SYSTEM'S FIRST PRE-REGISTERED CALL. Three blocks, each with its grading rule; outcomes enter this record after the 2026-09-30 print (release + 10-K). Nothing here scores until graded. Outcomes are graded after the print and enter the same ledger.
Management guidance [I]
claim Murphy-bias-implied FQ4 FY2026 revenue $57.4–60.1B vs the $49.0–51.0B guide
arithmetic guide $50.0B ± $1.0B (G-MU-Q3FY2026-01); all-claims mean bias +14.8% (n = 16 graded range claims, ONE REGIME — rising HBM demand, FQ4-25 → FQ3-26; 2 HIT / 14 MISS, every miss on the high side) → 50.0 × 1.148 = $57.4B; revenue-leg biases +9.1% / +27.6% / +23.7% → mean +20.1% → 50.0 × 1.201 = $60.1B
falsifier print ≤ $51.0B = regime-note sign-flip data point
grading rule R2 / R6: the guide grades HIT/MISS as stated ($49.0–51.0B); the implied range grades as an [I] — HIT if the printed revenue lands in $57.4–60.1B, MISS otherwise, with the signed bias recorded; a print ≤ $51.0B additionally logs "regime-note sign-flip data point #1" on voices/management/MU/murphy-mark.json
The ladder read [I]
claim the first US-side memory read — MU's own 10-K (FY2026, filed after the 2026-09-30 release) fills memory_bandwidth cells for MU; expected states below, each graded HIT/MISS on the read
gm_pricing_powerPRESENT — decelerating (ASP rising, rate lower than the prior report; the peer decelerated mid-60% → mid-30%)capacity_reservationABSENT (no customer reserving MU capacity in MU's own words; if the $100B RPO reads as reserved capacity in [F], this expectation MISSES and the bottleneck factor moves)prepaymentsUNKNOWN → to be read (no expectation registered; the contract-liability line is read cold)lead_timeWEAK-or-ABSENT (supply-constraint prose without a number, or nothing)take_or_payABSENT (as seller; MU's own purchase commitments are the other direction)
grading rule STATE-MATCH: each expectation is HIT if the read's admitted state equals the expected state (PRESENT-decelerating counts HIT only if the ASP sentence shows a lower rate than the prior report; WEAK-or-ABSENT counts HIT on either), MISS otherwise; prepayments is not graded (no expectation). The read is the A-MEMORY expresser note's next step ("MU (owned) reports 2026-09-30 and is the next expresser").
The scoring rep [I]
claim NOTHING in the print raises MU's conviction (the binder is the admitted-voice count, 3.75 → 3); a gross margin below 84.6% OR revenue ≤ the $51.0B guide LOWERS it through valuation (5 → 4) and, on a GM decline, arms the live falsifier's first trigger
falsifier a post-print rep that raises conviction without a new admitted voice naming MU — that would mean the binder was misidentified
grading rule HIT if the post-print rep's conviction is ≤ 3 absent a new admitted voice; MISS if it is raised without one. The lowering leg grades on its own condition (GM < 84.6% or revenue ≤ $51.0B → valuation ceiling ≤ 4 at the post-print rep).
Read from ledger/performance-ledger.json (record MU, preregistered_20260930).