docs(psychometrics): bound routing score interpretation - #1062

Draft

seonghobae wants to merge 29 commits into

ContextualWisdomLab:mainfrom

seonghobae:codex/autoresearch-psychometric-validity-20260904

Commits on Sep 4, 2026

experiment: index psychometric contexts once
seonghobae
committed
0561c9b
docs(routing): define psychometric accuracy-time KPI
seonghobae
committed
b2f9011
experiment: trim psychometric rows without full scans
seonghobae
committed
44acd6a
docs(routing): record observation latency KPI
seonghobae
committed
d213489
experiment: measure held-out psychometric routing
seonghobae
committed
0ae0ed8
experiment: interpolate psychometric warm starts
seonghobae
committed
50d91c9
fix(routing): retain iterable candidates during interpolation
seonghobae
committed
f5abcf0
docs(routing): record held-out accuracy evidence
seonghobae
committed
dc638ac
docs(adr): govern psychometric warm starts
seonghobae
committed
11de547
docs(research): complete warm-start attribution
seonghobae
committed
f9bb820
fix(routing): reject opposite psychometric contexts
seonghobae
committed
c22c08e
fix(routing): reject non-finite psychometric vectors
seonghobae
committed
473b170
docs(adr): refresh psychometric guard evidence
seonghobae
committed
90b4013
fix(benchmark): require psychometric Python runtime
seonghobae
committed
3c22bf3
Merge commit '3c22bf31c33520b128cb2c449d03e3762e1f4c75' into codex/autoresearch-psychometric-observe-20260904
seonghobae
committed
99efa19
Merge commit '99efa19e' into codex/autoresearch-psychometric-heldout-20260904
seonghobae
committed
ab172f7
docs(adr): refresh stacked runtime evidence
seonghobae
committed
cbe12d4
fix(benchmark): report expected Brier score
seonghobae
committed
c5a0963
docs(psychometrics): bound routing score interpretation
seonghobae
committed
9fb481b
test(research): enforce paper inventory coverage
seonghobae
committed
eef69f7
test(research): cover publisher paper identifiers
seonghobae
committed
23e9ee1
fix(routing): gate unvalidated psychometric evidence
seonghobae
committed
c3feada
docs(changelog): record psychometric routing gates
seonghobae
committed
740457c
fix(routing): discard stale psychometric evidence
seonghobae
committed
174bdfa
fix(routing): purge invalid restored evidence
seonghobae
committed
fd536bb
fix(routing): bind evidence to decode policy
seonghobae
committed
0554b3a
docs(routing): record psychometric policy boundary
seonghobae
committed
61bb6df
fix(evidence): serialize psychometric persistence
seonghobae
committed
ffb1383
docs(evidence): record persistence concurrency repair
seonghobae
committed
cafdc7c

Navigation Menu

Uh oh!

docs(psychometrics): bound routing score interpretation - #1062

docs(psychometrics): bound routing score interpretation#1062
seonghobae wants to merge 29 commits into
ContextualWisdomLab:mainfrom
seonghobae:codex/autoresearch-psychometric-validity-20260904

Commits on Sep 4, 2026

experiment: index psychometric contexts once

docs(routing): define psychometric accuracy-time KPI

experiment: trim psychometric rows without full scans

docs(routing): record observation latency KPI

experiment: measure held-out psychometric routing

experiment: interpolate psychometric warm starts

fix(routing): retain iterable candidates during interpolation

docs(routing): record held-out accuracy evidence

docs(adr): govern psychometric warm starts

docs(research): complete warm-start attribution

fix(routing): reject opposite psychometric contexts

fix(routing): reject non-finite psychometric vectors

docs(adr): refresh psychometric guard evidence

fix(benchmark): require psychometric Python runtime

Merge commit '3c22bf31c33520b128cb2c449d03e3762e1f4c75' into codex/autoresearch-psychometric-observe-20260904

Merge commit '99efa19e' into codex/autoresearch-psychometric-heldout-20260904

docs(adr): refresh stacked runtime evidence

fix(benchmark): report expected Brier score

docs(psychometrics): bound routing score interpretation

test(research): enforce paper inventory coverage

test(research): cover publisher paper identifiers

fix(routing): gate unvalidated psychometric evidence

docs(changelog): record psychometric routing gates

fix(routing): discard stale psychometric evidence

fix(routing): purge invalid restored evidence

fix(routing): bind evidence to decode policy

docs(routing): record psychometric policy boundary

fix(evidence): serialize psychometric persistence

docs(evidence): record persistence concurrency repair