Built independently by an author, for readers. Read the story and support ChapterPal
ChapterPal Logo
Log inSign up

keyword

feedback learning

Feedback learning is a machine learning paradigm in which an artificial intelligence model refines its behavior, outputs, or internal parameters using evaluative responses and guidance received after generating predictions or actions. Rather than relying exclusively on fixed, pre-labeled training pairs as in standard supervised learning, this approach incorporates ongoing signals such as human preference rankings, scalar ratings, error corrections, or natural language critiques to steer the system. Primarily used in the training and alignment of language models, feedback learning enables systems to better accommodate nuanced objectives, subjective human values, and safety standards through methods like reinforcement learning from feedback and direct preference optimization, thereby closing the gap between raw generative capability and human-aligned performance.

1 item

The Past, Present and Better Future of Feedback Learning in Large Language Models for Subjective Human Preferences and Values

The Past, Present and Better Future of Feedback Learning in Large Language Models for Subjective Human Preferences and Values

Hannah Kirk, Andrew M. Bean, Bertie Vidgen, Paul Röttger, Scott Hale

OrganizationsBocconi UniversityMeedanUniversity of Oxford

Why you should read this

Systematizes the evolution of human feedback learning across 95 studies to identify critical conceptual and practical challenges in aligning language models with subjective preferences and values.

Human feedback is increasingly used to steer the behaviours of Large Language Models (LLMs). However, it is unclear how to collect and incorporate feedback in a way that is efficient, effective and unbiased, especially for highly subjective human preferences and values. In this paper, we survey existing approaches for learning from human feedback, drawing on 95 papers primarily from the ACL and arXiv repositories. First, we summarise the past, pre-LLM trends for integrating human feedback into language models. Second, we give an overview of present techniques and practices, as well as the motivations for using feedback; conceptual frameworks for defining values and preferences; and how feedback is collected and from whom. Finally, we encourage a better future of feedback learning in LLMs by raising five unresolved conceptual and practical challenges.

Added

2026-09-26