Universal Self-Consistency

Sample Count Is Not Enough: Candidate-Generation Strategy Shapes the Energy and Performance of LLM Test-Time Scaling
arXiv:2609.19499 — For self-consistency, RASC, ESC, and best-of-N, candidate count N is not enough. Batching the same N can change A100 energy and latency by several times.
Mobina Kashaniyan, Ali Jannesari
Posted Nov 1, 2026 Conference Paper, Workshop Paper, Preprint, Large Language Models, High-Performance Computing, Energy Efficiency
PDF DOI PDF arXiv DOI Google Scholar Research Summary Local PDF
Sample Count Is Not Enough (arXiv:2609.19499): Why Candidate-Generation Strategy Matters for LLM Test-Time Scaling Energy and Performance
arXiv:2609.19499 — For self-consistency, RASC, ESC, Universal Self-Consistency, and best-of-N, sample count N is not enough. At N=8, 1×8 vs 8×1 changes A100 energy by about 4.6–4.9×.
Mobina Kashaniyan
Posted Sep 16, 2026 Last updated Sep 17, 2026 Large Language Models, High-Performance Computing, Energy Efficiency, Test-Time Scaling, Research Summary

© 2026 Mobina Kashaniyan.