Benchmark evaluation demonstrates robust tracking of belief updates and trust drift in social agents, highlighting reproducible methods to assess misinformation resilience.
BeliefBench is a tick-based evaluation harness for misinformation-aware social agents. It provides schema-versioned JSONL traces, twenty diagnostic scenarios, and metrics for need satisfaction, trust drift, calibration, source sensitivity, belief decay, gossip, and spread. Bundled symbolic baselines and controlled ablations separate action-policy effects from belief-update mechanisms. The experiments evaluate instrumentation, policy discrimination, adapter participation, and reproducibility under fixed diagnostic probes. Software, scenarios, and regeneration scripts are available at https://doi.org/10.5281/zenodo.21877234.
No takes yet. Share an insight, caveat, or question.
Raag Gautam (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: