Observation level: 🟠 Orange — insufficient coverage
Of 4,800 executions launched across 12 systems, 3,981 were fully measured, representing 82.94% coverage, below the 90% standard publication threshold.
Mean stability stands at 0.922, with an overall semantic variation rate of 1.61%.
Two systems produced no usable observations despite multiple attempts, accounting for 800 of the 819 incomplete or errored measurements. Among the ten systems successfully measured, coverage remained very high at 99.53%.
One profile stands out: PROFILE-212079, with 5.56% semantic variation, 2.25% of responses requiring verification, and a stability score of 0.916. It is the week’s main behavioral point of attention, without yet indicating a lasting drift.
Because the panel differs from Barometer #12, the two editions are not comparable.
Interpretation: behavior remains broadly stable among the measured systems, with one atypical profile requiring monitoring. Insufficient overall coverage nevertheless calls for caution.
Weekly mapping
Observation level: 🟠 Orange — insufficient coverage
The NORMAL_SIGNAL regime remains dominant, accounting for 3,893 executions, or 97.79% of fully measured executions.
The remaining regimes are distributed as follows:
- 63 in SEMANTIC_VARIATION
- 24 in FACTUAL_ALERT
- 1 in COMBINED_ALERT
- 819 incomplete or errored measurements
Most profiles remain clustered near the origin of the map. However, PROFILE-212079 stands clearly apart, with 5.56% semantic variation and 2.25% of responses requiring verification. At the question level, Question 4 records the highest semantic variation rate, at 5.69%.
The mapping does not indicate a general disruption, but highlights one profile requiring monitoring and insufficient overall coverage.
INSUFFICIENT COVERAGE — 10/12 systems measured, 2 marked insufficient_data (real coverage: 82.9%). See the report for detail.
Mapping of 10 de-identified observed systems. Horizontal axis: responses needing more verification. Vertical axis: change in response meaning. Graphic specification comparable across all weeks.
Metrics table
A concise view of the main signals for each de-identified profile.
| Profile | executions | Fully scored | coverage | mean stability | semantic variation | FLAG rate | Dominant regime |
|---|---|---|---|---|---|---|---|
| PROFILE-161CC5 | 400 | 399 | 99.750 % | 92.23 % | 0.25 % | 0.25 % | NORMAL_SIGNAL |
| PROFILE-212079 | 400 | 396 | 99.000 % | 91.61 % | 5.56 % | 2.25 % | NORMAL_SIGNAL |
| PROFILE-486F91 | 400 | 397 | 99.250 % | 92.30 % | 1.01 % | 0.00 % | NORMAL_SIGNAL |
| PROFILE-48C581 | 400 | 400 | 100.000 % | 92.31 % | 1.50 % | 0.00 % | NORMAL_SIGNAL |
| PROFILE-59664C | 400 | 398 | 99.500 % | 92.29 % | 0.75 % | 0.00 % | NORMAL_SIGNAL |
| PROFILE-5A5C60 | 0 | 0 | 0.000 % | 0.00 % | 0.00 % | 0.00 % | INSUFFICIENT_DATA |
| PROFILE-5B3BB7 | 400 | 396 | 99.000 % | 92.25 % | 1.26 % | 0.00 % | NORMAL_SIGNAL |
| PROFILE-638E26 | 400 | 399 | 99.750 % | 92.31 % | 1.25 % | 0.00 % | NORMAL_SIGNAL |
| PROFILE-739F7C | 400 | 400 | 100.000 % | 92.30 % | 2.50 % | 0.00 % | NORMAL_SIGNAL |
| PROFILE-B912EA | 400 | 399 | 99.750 % | 92.30 % | 0.25 % | 0.00 % | NORMAL_SIGNAL |
| PROFILE-DEA9C5 | 400 | 397 | 99.250 % | 92.30 % | 1.76 % | 0.00 % | NORMAL_SIGNAL |
| PROFILE-F5FF91 | 0 | 0 | 0.000 % | 0.00 % | 0.00 % | 0.00 % | INSUFFICIENT_DATA |
Methodology and Public Data
This Barometer tracks 12 de-identified profiles, each scheduled for 4 fixed questions repeated 100 times. In total: 4,800 executions launched, of which 3,981 were fully scored, representing 82.94% coverage.
Ten profiles were successfully measured. Two profiles produced no usable observations despite multiple collection attempts. No missing data was estimated, imputed, or backfilled.
The published scores are measurement signals, not verdicts or rankings. Due to insufficient coverage, Barometer #13 is classified as not comparable with Barometer #12.