Directional Preference Alignment and RLHF Pass@k Performance on HumanEval at Scale

Abstract

Abstract: This report synthesises findings from 10 peer-reviewed papers addressing the following research question: How does the pass@k metric for Directional Preference Alignment compare to RLHF on the HumanEval benchmark when scaling model parameters from 13B to 175B. We introduce ChatGLM, an evolving family of large language models that we have been developing over time. This report primarily focuses on the GLM-4 language series, which includes GLM-4, GLM-4-Air, and GLM-4-9B. 10 claims were extracted from source literature; 9 were independently verified against retrieved documents. An automated multi-reviewer quality assessment produced a score of 8.5/10. This report is a machine-generated literature synthesis and does not constitute original research.

Research Question

How does the pass@k metric for Directional Preference Alignment compare to RLHF on the HumanEval benchmark when scaling model parameters from 13B to 175B?

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