# Deep Reinforcement Learning from Policy-Dependent Human Feedback

arXiv: https://arxiv.org/abs/1902.04257 · HTML: https://mlsanity.com/paper/1902.04257/

## Key facts

- Cited by 56 papers in the map (data through 14 Oct 2026).
- Placed in Reward modeling and RLHF, one of 10 directions in LLM alignment / RLHF / preference optimisation.
- Builds on 12 of its 41 references that are in the map.
- First posted on arXiv 12 Feb 2019.

## Placed in

- LLM alignment / RLHF / preference optimisation / [Reward modeling and RLHF](https://mlsanity.com/alignment/reward-modeling-and-rlhf/)

## Builds on (most cited first, 12 of 12)

- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [Deep reinforcement learning from human preferences](https://mlsanity.com/paper/1706.03741/) (2017)
- [Continuous control with deep reinforcement learning](https://mlsanity.com/paper/1509.02971/) (2015)
- [Asynchronous Methods for Deep Reinforcement Learning](https://mlsanity.com/paper/1602.01783/) (2016)
- [Rainbow: Combining Improvements in Deep Reinforcement Learning](https://mlsanity.com/paper/1710.02298/) (2017)
- [Guided Cost Learning: Deep Inverse Optimal Control via Policy Optimization](https://mlsanity.com/paper/1603.00448/) (2016)
- [Interactive Learning from Policy-Dependent Human Feedback](https://mlsanity.com/paper/1701.06049/) (2017)
- [Deep TAMER: Interactive Agent Shaping in High-Dimensional State Spaces](https://mlsanity.com/paper/1709.10163/) (2017)
- [A Deep Hierarchical Approach to Lifelong Learning in Minecraft](https://mlsanity.com/paper/1604.07255/) (2016)
- [Environment-Independent Task Specifications via GLTL](https://mlsanity.com/paper/1704.04341/) (2017)
- [Control of Memory, Active Perception, and Action in Minecraft](https://mlsanity.com/paper/1605.09128/) (2016)

## Built on by (most cited first, 30 of 56)

- [Open Problems and Fundamental Limitations of Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2307.15217/) (2023)
- [Video PreTraining (VPT): Learning to Act by Watching Unlabeled Online Videos](https://mlsanity.com/paper/2206.11795/) (2022)
- [PEBBLE: Feedback-Efficient Interactive Reinforcement Learning via Relabeling Experience and Unsupervised Pre-training](https://mlsanity.com/paper/2106.05091/) (2021)
- [A Survey of Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2312.14925/) (2023)
- [Vision-Language Models as Success Detectors](https://mlsanity.com/paper/2303.07280/) (2023)
- [B-Pref: Benchmarking Preference-Based Reinforcement Learning](https://mlsanity.com/paper/2111.03026/) (2021)
- [Motif: Intrinsic Motivation from Artificial Intelligence Feedback](https://mlsanity.com/paper/2310.00166/) (2023)
- [Reinforcement Learning Optimization for Large-Scale Learning: An Efficient and User-Friendly Scaling Library](https://mlsanity.com/paper/2506.06122/) (2025)
- [Imitation Learning: Progress, Taxonomies and Challenges](https://mlsanity.com/paper/2106.12177/) (2021)
- [A Comprehensive Survey of Foundation Models in Medicine](https://mlsanity.com/paper/2406.10729/) (2024)
- [Efficient Learning of Safe Driving Policy via Human-AI Copilot Optimization](https://mlsanity.com/paper/2202.10341/) (2022)
- [Skill Preferences: Learning to Extract and Execute Robotic Skills from Human Feedback](https://mlsanity.com/paper/2108.05382/) (2021)
- [Beyond Reward: Offline Preference-guided Policy Optimization](https://mlsanity.com/paper/2305.16217/) (2023)
- [MCU: An Evaluation Framework for Open-Ended Game Agents](https://mlsanity.com/paper/2310.08367/) (2023)
- [Cognitive science as a source of forward and inverse models of human decisions for robotics and control](https://mlsanity.com/paper/2109.00127/) (2021)
- [Continual Learning for Instruction Following from Realtime Feedback](https://mlsanity.com/paper/2212.09710/) (2022)
- [Accelerating Reinforcement Learning of Robotic Manipulations via Feedback from Large Language Models](https://mlsanity.com/paper/2311.02379/) (2023)
- [Uni-RLHF: Universal Platform and Benchmark Suite for Reinforcement Learning with Diverse Human Feedback](https://mlsanity.com/paper/2402.02423/) (2024)
- [MAP: Multi-Human-Value Alignment Palette](https://mlsanity.com/paper/2410.19198/) (2024)
- [Continual Learning for Grounded Instruction Generation by Observing Human Following Behavior](https://mlsanity.com/paper/2108.04812/) (2021)
- [Text2Chart31: Instruction Tuning for Chart Generation with Automatic Feedback](https://mlsanity.com/paper/2410.04064/) (2024)
- [Primitive Skill-based Robot Learning from Human Evaluative Feedback](https://mlsanity.com/paper/2307.15801/) (2023)
- [Enhancing Rating-Based Reinforcement Learning to Effectively Leverage Feedback from Large Vision-Language Models](https://mlsanity.com/paper/2506.12822/) (2025)
- [Recent Advances in Leveraging Human Guidance for Sequential Decision-Making Tasks](https://mlsanity.com/paper/2107.05825/) (2021)
- [Active Reward Learning from Multiple Teachers](https://mlsanity.com/paper/2303.00894/) (2023)
- [CREW: Facilitating Human-AI Teaming Research](https://mlsanity.com/paper/2408.00170/) (2024)
- [Sample-Efficient Preference-based Reinforcement Learning with Dynamics Aware Rewards](https://mlsanity.com/paper/2402.17975/) (2024)
- [ASHA: Assistive Teleoperation via Human-in-the-Loop Reinforcement Learning](https://mlsanity.com/paper/2202.02465/) (2022)
- [Shattering the Agent-Environment Interface for Fine-Tuning Inclusive Language Models](https://mlsanity.com/paper/2305.11455/) (2023)
- [BEDD: The MineRL BASALT Evaluation and Demonstrations Dataset for Training and Benchmarking Agents that Solve Fuzzy Tasks](https://mlsanity.com/paper/2312.02405/) (2023)

## Abstract (arXiv)

To widen their accessibility and increase their utility, intelligent agents must be able to learn complex behaviors as specified by (non-expert) human users. Moreover, they will need to learn these behaviors within a reasonable amount of time while efficiently leveraging the sparse feedback a human trainer is capable of providing. Recent work has shown that human feedback can be characterized as a critique of an agent's current behavior rather than as an alternative reward signal to be maximized, culminating in the COnvergent Actor-Critic by Humans (COACH) algorithm for making direct policy updates based on human feedback. Our work builds on COACH, moving to a setting where the agent's policy is represented by a deep neural network. We employ a series of modifications on top of the original COACH algorithm that are critical for successfully learning behaviors from high-dimensional observations, while also satisfying the constraint of obtaining reduced sample complexity. We demonstrate the effectiveness of our Deep COACH algorithm in the rich 3D world of Minecraft with an agent that learns to complete tasks by mapping from raw pixels to actions using only real-time human feedback in 10-15 minutes of interaction.