AI truth finding improves when accounting for shared errors among agents
Where Is the Truth When AI Agents Disagree? Dependence-Aware Belief Propagation for Unsupervised Truth Discovery (Extended Version)
Artificial IntelligenceComputation and LanguageDatabases
Summary
When many AI programs answer the same question, they often disagree and sometimes make similar mistakes if based on the same underlying model. The authors study how to figure out the true answer and which AI agents are more reliable without access to correct labels. They introduce a method called BP-TD that adjusts for groups of similar agents so they don't unfairly sway the consensus. Tested on a benchmark with many questions, BP-TD outperformed simple majority voting and better ranked the agents by accuracy.
What this means in practice
- •For ai platform developers: Improve ensemble predictions by accounting for shared biases among similar AI models to better infer correct answers without labeled data.
- •For crowdsourcing project managers: More accurately aggregate multiple annotators' answers by adjusting for dependencies when workers share common error sources, improving label quality without gold data.
Authors
Laure Berti-Equille
Abstract
LLM agents often give conflicting answers to the same question, and many deployed agents are variants of one base model, so their errors are correlated. We study truth discovery among such agents: inferring the true answer and each agent's reliability without labels. Classical truth-discovery and crowdsourcing estimators reward an agreeing bloc and can fall below majority vote. We propose BP-TD, which adds per-agent decorrelation weights to mean-field variational inference on a one-coin Dawid-Skene model, so that agents sharing a base model count together as about one vote. Its first iteration is exactly a weighted vote; later iterations estimate agent reliability. With 13 local agents on MMLU, BP-TD beats majority vote without any labels (0.716 vs. 0.624 on 160 questions; 0.707 vs. 0.674 on 10,000) and ranks agents by their true accuracy, which classical estimators invert. The gain depends on the pool: over 2,852 sub-pools of four benchmarks, decorrelation helps when the largest correlated bloc is weaker than the other agents and hurts when it is the strongest, and we found no label-free statistic that tells the two cases apart. Labels change the picture quickly: with 20 labelled questions, selecting the single best agent is more accurate than every aggregator we tested on MMLU, ARC and TruthfulQA.