Cognition · Software component
Majority Vote Aggregator
Software componentCognitionCognition & Memoryarc:MajorityVoteAggregator
A self-consistency aggregator that selects the answer appearing most frequently across sampled reasoning paths, counting every path equally.
Responsibility. Selects the most frequent answer across equally weighted samples.
Also known as: Standard Self-Consistency voting, Unweighted majority vote
Variant of Self-Consistency Aggregator abstract
When to choose. Choose for easy problems or small sample counts (k=3-5) where vote distributions are already clear and quality-assessment overhead is not justified.
Relationships
alternative to variability
Design guidance
- SHOULD NOT be relied on when errors are systematic or correlated across samples, since all paths then converge confidently on the wrong answer.
Quantitative guidance
As stated by the sources; verify before use.
- GSM8K: ~58% (CoT) to 74% (k=5, T=0.7); gains of +17.9% GSM8K, +11.0% SVAMP, +3.9% ARC-challenge, +1.2% TriviaQA at k=20 (Ch5.3).
Classification
- Patterns
- Majority votingPositional (ranked-list) voting
- Quality attributes
- Maintainability (ISO/IEC 25010)Functional suitability: correctness and validity (ISO/IEC 25010 | NIST AI RMF: valid)
- Risks mitigated
- Path-specific calculation errors
Sources
- Ch5.3: T. Nguyen, "Self-Consistency Fundamentals," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 5.3. ISBN: 9798244538229.