ABC Align: Large Language Model Alignment for Safety & Accuracy

TL;DR

ABC Align introduces a new methodology for aligning large language models with organizational standards and preferences, enhancing safety and accuracy while maintaining reasoning abilities through innovative data and optimization techniques.

Contribution

It presents a novel alignment approach that integrates organizational standards into LLMs using synthetic data, preference optimization, and model quantization.

Findings

Mitigates bias in LLMs

Improves accuracy on standard benchmarks

Preserves reasoning capabilities

Abstract

Alignment of Large Language Models (LLMs) remains an unsolved problem. Human preferences are highly distributed and can be captured at multiple levels of abstraction, from the individual to diverse populations. Organisational preferences, represented by standards and principles, are defined to mitigate reputational risk or meet legislative obligations. In this paper, we present ABC Align, a novel alignment methodology for LLMs that enables integration of the standards and preferences of a large media organisation into the LLM itself. We combine a set of data and methods that build on recent breakthroughs in synthetic data generation, preference optimisation, and post-training model quantisation. Our unified approach mitigates bias and improves accuracy, while preserving reasoning capability, as measured against standard benchmarks.