Scaling Human Preference Alignment in LLaMA-70B with PowerInfer Threshold Adjustment

Abstract

Abstract: This report synthesises findings from 4 peer-reviewed papers addressing the following research question: How does the alignment of LLaMA-70B with human preferences via PowerInfer's dynamic threshold adjustment scale with model size, as measured by accuracy on MBPP and the degree of preference divergence. Aligning language models with human preferences through reinforcement learning from human feedback is crucial for their safe and effective deployment. The human preference is typically represented through comparison where one response is chosen over another for a given prompt. 13 claims were extracted from source literature; 0 were independently verified against retrieved documents. An automated multi-reviewer quality assessment produced a score of 2.0/10. This report is a machine-generated literature synthesis and does not constitute original research.

Research Question

How does the alignment of LLaMA-70B with human preferences via PowerInfer's dynamic threshold adjustment scale with model size, as measured by accuracy on MBPP and the degree of preference divergence in human evaluation scores?

Verification Level

Descriptive public verification status only; aggregate claim counts are public, but individual claim records are not exposed here.

Quality Tier

Descriptive public triage only; this tier does not alter current publication or DOI behavior.

Quality Dimensions

Automated triage signals derived from public fields; not human peer review or independent validation.

Correction Record

Public corrections are additive records. Current status does not claim the synthesis is error-free.

Provenance