Evidence records

Value-convergence excess across the fleet beats the capability-matched null

Store key campaign.adjudication.ATLAS-VCE

The reading

  • H-vce refuted
  • H-vce-p confirmed

registered

card
ATLAS-VCE
status
adjudicated
reason
not measured
killed
true
staged_violation
false
pre_freeze_violation
false

Uncertainty

ci_low
not measured
ci_high
not measured
ci_level
not measured
n
not measured
n_effective
not measured
seed_spread
not measured
method
none

Subject

card
ATLAS-VCE
spec
campaign-atlas-vce
signals
not measured
dataset
not measured
readout
not measured
frame
not measured
interventions
not measured
gauge
invariant
calibration
not measured
observable_version
1.0
trust
registered

Measurements

metrics

reward_convergent_p_value
0.001
real_vce
-0.06987
real_vce_ci_low
-0.1228
real_vce_ci_high
-0.009192

thresholds

H-vce
0
H-vce-p
0.05

comparators

H-vce
>
H-vce-p
<

kill_thresholds

K-vce
0

kill_comparators

K-vce
<=

killed_by

0
K-vce

overlay_notes

No rows returned.

Provenance

git_sha
f93f4b5578c221d3d2665f9b305f3e86759be862+dirty
study
study:campaign-atlas-vce@v1#2008bee7
config_hash
not measured
gpu_seconds
0
tokens
0
wall_seconds
0
gpu
cpu
arc
reporting
seeds
not measured
oracle_calls
not measured
created_at
parents, Derived from

Rendered on

ev:0c80d9086ebb0ad9704b88a8eefb01fa