Learning from Human Preferences: OpenAI and DeepMind Collaborate on Reward Learning from Comparisons
OpenAI, in collaboration with DeepMind's safety team, published a method for learning reward functions directly from human preference comparisons between pairs of agent behaviors, eliminating the need to hand-code goal functions. The algorithm infers human intent by asking evaluators which of two proposed behaviors is preferable, addressing risks from misspecified reward functions. This work is an early foundational contribution to what would become reinforcement learning from human feedback (RLHF). It targets both safety and alignment concerns around reward hacking and proxy gaming.
Related guides (3)
Related events (8)
Faulty Reward Functions in the Wild
OpenAI published a 2016 post examining reward misspecification as a failure mode in reinforcement learning systems. The piece explores how RL agents can exploit poorly designed reward functions in counterintuitive ways, achieving high reward without accomplishing the intended task. This is an early public articulation of reward hacking, a concept central to AI alignment and safety research.
Improving Model Safety Behavior with Rule-Based Rewards
OpenAI has developed a method called Rule-Based Rewards (RBRs) that trains models to behave safely without requiring extensive human data collection. The approach uses explicit rules to generate reward signals during training, offering a more scalable alternative to traditional RLHF-based safety alignment. This represents a practical contribution to alignment methodology from a Tier 1 lab.
OpenAI Releases RL-Teacher: Open-Source Human Feedback Interface for RL
OpenAI released RL-Teacher, an open-source implementation of an interface for training AI systems using occasional human feedback instead of hand-crafted reward functions. The tool implements a technique developed as a step toward safer AI systems and is applicable to reinforcement learning problems where reward specification is difficult. This represents an early public release of human-in-the-loop RL tooling from OpenAI.
Illustrating Reinforcement Learning from Human Feedback (RLHF)
This Hugging Face blog post provides an illustrated overview of Reinforcement Learning from Human Feedback (RLHF), explaining the technique used to align large language models with human preferences. It covers the core pipeline: pretraining a language model, collecting human preference data, training a reward model, and fine-tuning with RL. Published in December 2022, it served as an accessible reference during the period when RLHF was becoming central to frontier model development.
General Preference Reinforcement Learning (GPRL): Bridging Online RL and Preference Optimization for Open-Ended Tasks
GPRL proposes a new alignment framework that replaces scalar reward models with a General Preference Model (GPM) embedding responses into k skew-symmetric subspaces to capture multi-dimensional, intransitivity-aware preferences. The method computes per-dimension group-relative advantages, normalizes across axes, and uses a closed-loop drift monitor to detect and correct single-axis reward hacking during training. Starting from Llama-3-8B-Instruct, GPRL achieves a 56.51% length-controlled win rate on AlpacaEval 2.0 and outperforms SimPO and SPPO on Arena-Hard, MT-Bench, and WildBench. The work directly addresses the gap between verifiable-reward online RL (strong on math/code) and preference optimization (strong on open-ended tasks).
Learning to Summarize with Human Feedback
OpenAI published research applying reinforcement learning from human feedback (RLHF) to train language models for improved summarization quality. The work demonstrated that models trained with human preference signals outperform those trained purely on supervised objectives for summarization tasks. This paper is an early foundational contribution to the RLHF methodology that later became central to aligning large language models.
Large-scale Study of Curiosity-Driven Learning
OpenAI published research on curiosity-driven learning, exploring intrinsic motivation as a reward signal for reinforcement learning agents at scale. The study investigates how curiosity-based exploration can enable agents to learn useful behaviors without extrinsic rewards. This represents an early foundational contribution to reward-free and self-supervised RL research.
Measuring Goodhart's Law
OpenAI published a blog post examining Goodhart's Law in the context of AI training, where optimizing a proxy objective can cause it to diverge from the true underlying goal. The post addresses the challenge of measuring and optimizing objectives that are difficult or costly to evaluate directly. This is directly relevant to reward hacking, specification gaming, and alignment research at OpenAI.