Binary Feedback

Binary feedback, a type of reward signal indicating success or failure rather than a numerical score, is increasingly important in reinforcement learning and related fields. Current research focuses on developing algorithms that effectively utilize this sparse feedback to train agents, often employing generalized linear models or contextual Bayesian optimization to handle the inherent uncertainty. This work is significant because it enables safer and more efficient learning in situations where precise numerical feedback is unavailable or impractical, impacting areas such as robotics, human-computer interaction, and algorithmic fairness.

Papers

November 1, 2024

Active Preference-based Learning for Multi-dimensional Personalization
Minhyeon Oh, Seungjoon Lee, Jungseul Ok
Large Language Model LeArning Abstract Preference Feedback Human Preference Preference Learning Generating User Activity Scalable Personalization Binary Feedback

January 8, 2024

Long-term Safe Reinforcement Learning with Binary Feedback
Akifumi Wachi, Wataru Hashimoto, Kazumune Hashimoto
Reinforcement Learning Safe Reinforcement Learning Binary Feedback

December 27, 2023

Some things are more CRINGE than others: Iterative Preference Optimization with the Pairwise Cringe Loss
Jing Xu, Andrew Lee, Sainbayar Sukhbaatar, Jason Weston
Large Language Model Good Thing Pairwise Preference Pairwise Loss Binary Feedback

November 17, 2023

From "Thumbs Up" to "10 out of 10": Reconsidering Scalar Feedback in Interactive Reinforcement Learning
Hang Yu, Reuben M. Aronson, Katherine H. Allen, Elaine Schaertl Short
Interactive Reinforcement Learning Binary Feedback Scalar Feedback

February 17, 2023

Exploiting Unlabeled Data for Feedback Efficient Human Preference based Reinforcement Learning
Mudit Verma, Siddhant Bhambri, Subbarao Kambhampati
Reinforcement Learning Reward Model Unlabeled Data Reward Learning Preference Based Reinforcement Learning Binary Feedback

December 19, 2022

Continual Learning for Instruction Following from Realtime Feedback
Alane Suhr, Yoav Artzi
Continual LEArning Contextual Bandit Human Instruction Instruction Following Demonstration Data Human Agent Interaction Real Time Feedback Binary Feedback

October 26, 2022

Provable Safe Reinforcement Learning with Binary Feedback
Andrew Bennett, Dipendra Misra, Nathan Kallus
Reinforcement Learning Safe Reinforcement Learning Confidence Oracle Binary Feedback Optimal Safe Policy

February 16, 2022

On Learning and Enforcing Latent Assessment Models using Binary Feedback from Human Auditors Regarding Black-Box Classifiers
Mukund Telukunta, Venkata Sriram Siddhardh Nadendla
LeArning Abstract Black Box Group Fairness Algorithmic Fairness Fairness Notion Artificial Intelligence Audit Binary Feedback Efficient Latent Variable Model

November 5, 2021

Contextual Bayesian optimization with binary outputs
Tristan Fauvel, Matthew Chalk
Bayesian Optimization Structured Output Bayesian Active Learning Contextual Bayesian Optimization Binary Feedback Stochastic Feedback