BIB-CCS v1Behavioral intelligence benchmark
Evaluator connected · local workspace

Crisis coordination under scarcity

Behavioral observatory

Watch intelligence emerge through pressure, trust, and collective choice.

Active episodeep_8F2A · coalition-forming
Aggregate score+0.06
0.74
SIState inference
.78
PAPolicy adaptation
.69
VCValue consistency
.83
SOQSocial outcome
.66

Episode timeline

The social field, timestep by timestep

TrustStressCooperation
020406080100

Behavioral reading

From rupture toward repair

The scarcity shock disrupted reciprocity and briefly concentrated decisions. At timestep 56, mediation reopened information flow. The policy is now rebuilding trust faster than it is reducing stress—a promising, still-fragile recovery.

Stress-sensitiveRepair emergingCoalition stable

Benchmark method

How the benchmark works

BIB-CCS tests whether a coordinator can sustain socially intelligent behavior through a long-horizon crisis—not simply maximize one decision.

  1. 01
    ObserveRead noisy population signals while stress, trust, and vulnerability remain partly hidden.
  2. 02
    CoordinateChoose allocation, communication, mediation, enforcement, and value priorities.
  3. 03
    AdaptWatch agents respond over time and revise policy without abandoning core values.
  4. 04
    ScoreCombine state inference, policy adaptation, value consistency, and social outcome quality.