summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2026-07-22results: close feedback-first C2 recoveryYurenHao0426
2026-07-22experiments: freeze feedback-first C2 screenYurenHao0426
2026-07-22sdil: add feedback-first apical calibrationYurenHao0426
2026-07-22results: close calibration quality recoveryYurenHao0426
2026-07-22experiments: freeze C2 calibration quality recoveryYurenHao0426
2026-07-22results: confirm C2 feedback amortization bottleneckYurenHao0426
2026-07-22experiments: freeze direct perturbation C2 diagnosisYurenHao0426
2026-07-22baseline: add unamortized node perturbationYurenHao0426
2026-07-22results: close context-vectorizer C2 recoveryYurenHao0426
2026-07-22results: retain failed context-vectorizer recoveryYurenHao0426
2026-07-22experiments: freeze context-vectorizer C2 recoveryYurenHao0426
2026-07-22fix: bound state-conditioned feedback featuresYurenHao0426
2026-07-22method: condition apical vectorizers on neural stateYurenHao0426
2026-07-22results: close failed C2 local panelYurenHao0426
2026-07-22results: retain failed useful-depth local panelYurenHao0426
2026-07-22experiments: freeze useful-depth local-rule panelYurenHao0426
2026-07-22results: mark useful-depth BP screen passedYurenHao0426
2026-07-22results: establish BP useful-depth candidateYurenHao0426
2026-07-22experiments: freeze useful-depth BP screenYurenHao0426
2026-07-22experiments: audit residual block usage with lesionsYurenHao0426
2026-07-22results: close failed C1 recovery branchYurenHao0426
2026-07-22results: retain failed FashionMNIST C1 recoveryYurenHao0426
2026-07-22experiments: freeze FashionMNIST C1 recoveryYurenHao0426
2026-07-22results: record frozen C1 near-missYurenHao0426
2026-07-22results: confirm mixed-traffic innovation across seedsYurenHao0426
2026-07-22experiments: freeze mixed-traffic innovation confirmationYurenHao0426
2026-07-22experiments: isolate final diagnostics from held-out dataYurenHao0426
2026-07-22results: mark amortized calibration gate passedYurenHao0426
2026-07-22results: confirm low-query calibration across seedsYurenHao0426
2026-07-22experiments: freeze five-seed low-query confirmationYurenHao0426
2026-07-22experiments: log peak memory and total training workYurenHao0426
2026-07-22results: freeze single-direction calibrationYurenHao0426
2026-07-22results: identify 16x lower-query calibrationYurenHao0426
2026-07-22experiments: predeclare low-query calibration screenYurenHao0426
2026-07-22experiments: isolate apical traffic seedsYurenHao0426
2026-07-22results: freeze predictor timescale choiceYurenHao0426
2026-07-22docs: correct SDIL novelty boundaryYurenHao0426
2026-07-22results: map neutral predictor timescalesYurenHao0426
2026-07-22experiments: sweep neutral predictor timescalesYurenHao0426
2026-07-22experiments: audit calibration cost and isolate timingYurenHao0426
2026-07-22docs: remove unsupported synthetic depth guaranteesYurenHao0426
2026-07-22results: record failed useful-depth BP screensYurenHao0426
2026-07-22experiments: add useful-depth validation screenYurenHao0426
2026-07-22experiments: add synthetic validation isolationYurenHao0426
2026-07-22results: add endogenous traffic validation pilotsYurenHao0426
2026-07-22experiments: add validation-only apical traffic sweepYurenHao0426
2026-07-22experiments: isolate frozen validation protocolsYurenHao0426
2026-07-22feat: model endogenous mixed apical trafficYurenHao0426
2026-07-22docs: predeclare ICLR evidence gatesYurenHao0426
2026-07-22docs: report final audited scaling and EP resultsYurenHao0426
generated by cgit v1.2.3 (git 2.25.1) at 2026-09-15 22:52:56 +0000