Learning Whom to Trust : Decision-Generated Credibility in Social Learning
2026-08-25 • Neural and Evolutionary Computing
Neural and Evolutionary Computing
AI summaryⓘ
The authors explore how people learning together can help or hurt the group's overall knowledge when individuals share information based on how confident they feel about their choices. They use a mathematical model where agents make decisions and express confidence, which influences how much others trust and learn from them. Their findings show that moderate sharing of confident opinions helps correct mistakes faster, but too much sharing can cause everyone to agree on wrong answers. They also find that confidence plays two roles: it can both spread errors when trusted too much, and help keep learning stable when combined with personal experience.
reinforcement learningdrift-diffusion processdecision confidencesocial learningcommunity couplingMonte Carlo simulationsinformation transmissionconsensus formationcross-community permeabilityerror amplification
Authors
Gabriel Bontemps, Abhishek Banerjee
Abstract
Social interaction can improve collective learning but also amplify early mistakes. We study this tension when the credibility of social information is generated by the sender's own decision process rather than fixed ex ante. Reinforcement-learning agents make binary choices through a drift--diffusion process that jointly determines choice, decision time, and confidence; decision confidence then becomes social credibility by weighting anticipatory influence and retrospective social learning. Under balanced community exposure, the anticipatory field admits an exact quotient representation. Its local Jacobian is a scalar decision-sensitivity term multiplying the community-coupling matrix, which yields a common-mode amplification threshold and an analytical role for cross-community permeability in damping relative community differences. Monte Carlo experiments show the corresponding non-monotone performance pattern: moderate transmission accelerates correction, whereas strong transmission can lock populations into wrong consensus; low permeability instead sustains disagreement. Ablations reveal a dual role for confidence: credibility-sensitive transmission amplifies social error, while confidence-dependent private learning stabilises it. The model yields testable predictions linking sender confidence to receiver behaviour conditional on accuracy.