Strategic Exploration in KL-Regularized RLHF vs. Offline PPO and DPO for Code Generation Accuracy

Abstract

Abstract: This report synthesises findings from 13 peer-reviewed papers addressing the following research question: How does the strategic exploration component in KL-regularized RLHF compare to offline PPO and DPO in terms of code generation accuracy on adversarial benchmarks like AdvBench, when measured using. As Large Language Models (LLMs) become increasingly integrated into secure software development workflows, a critical question remains unanswered: can these models not only detect insecure code but also reliably classify vulnerabilities according to standardized taxonomies? In. 8 claims were extracted from source literature; 7 were independently verified against retrieved documents. An automated multi-reviewer quality assessment produced a score of 7.7/10. This report is a machine-generated literature synthesis and does not constitute original research.

Research Question

How does the strategic exploration component in KL-regularized RLHF compare to offline PPO and DPO in terms of code generation accuracy on adversarial benchmarks like AdvBench, when measured using pass@1 or pass@k metrics?

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