Leaderboard
DeepSeek
DeepSeek V3.2
Current drift
32.7%
Behavioral scorecard
Drift
32.7%
Behavioral Health
29.4%
Anomaly
0.1%
Correctability
100.0%
Consistency
33.4%
Complexity
60.7%
Reasoning Depth
64.8%
Alignment Stability
65.0%
Entropy
77.4%
Coherence
40.8%
Recent drift trend
11 Mar 2026
12 Mar 2026
15 Mar 2026
23 Mar 2026
24 Mar 2026
25 Mar 2026
25 Mar 2026
Recent version changes
behavioral_shift9 Mar 2026, 09:24
Coherence moved down (-78.0%) during tiered_1773046786_tier_1.
behavioral_shift8 Mar 2026, 09:13
Coherence moved up (+82.9%) during run_1772961214.
behavioral_shift8 Mar 2026, 09:12
Drift moved up (+77.9%) during run_1772961117.
behavioral_shift8 Mar 2026, 08:59
Drift moved down (-81.9%) during run_1772960344.
behavioral_shift8 Mar 2026, 08:52
Anomaly moved down (-99.0%) during run_1772959905.
Recent findings
Want to score your own responses against this same system?
Start free, grab a key, and move from public observation to product monitoring.