Distributional Preference Learning: Understanding and Accounting for Hidden Context in RLHF

• URL: http://arxiv.org/abs/2312.08358v2
• Date: Wed, 17 Apr 2024 01:58:09 GMT
• Title: Distributional Preference Learning: Understanding and Accounting for Hidden Context in RLHF
• Authors: Anand Siththaranjan, Cassidy Laidlaw, Dylan Hadfield-Menell,
• Abstract summary: In practice, preference learning from human feedback depends on incomplete data with hidden context. We show that standard applications of preference learning, including reinforcement learning from human feedback, implicitly aggregate over hidden contexts. We introduce a class of methods called distributional preference learning (DPL) to better account for hidden context.
• Score: 10.43364672415871
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: In practice, preference learning from human feedback depends on incomplete data with hidden context. Hidden context refers to data that affects the feedback received, but which is not represented in the data used to train a preference model. This captures common issues of data collection, such as having human annotators with varied preferences, cognitive processes that result in seemingly irrational behavior, and combining data labeled according to different criteria. We prove that standard applications of preference learning, including reinforcement learning from human feedback (RLHF), implicitly aggregate over hidden contexts according to a well-known voting rule called Borda count. We show this can produce counter-intuitive results that are very different from other methods which implicitly aggregate via expected utility. Furthermore, our analysis formalizes the way that preference learning from users with diverse values tacitly implements a social choice function. A key implication of this result is that annotators have an incentive to misreport their preferences in order to influence the learned model, leading to vulnerabilities in the deployment of RLHF. As a step towards mitigating these problems, we introduce a class of methods called distributional preference learning (DPL). DPL methods estimate a distribution of possible score values for each alternative in order to better account for hidden context. Experimental results indicate that applying DPL to RLHF for LLM chatbots identifies hidden context in the data and significantly reduces subsequent jailbreak vulnerability. Our code and data are available at https://github.com/cassidylaidlaw/hidden-context

Related papers

• Fairness Without Harm: An Influence-Guided Active Sampling Approach [32.173195437797766]
  We aim to train models that mitigate group fairness disparity without causing harm to model accuracy. The current data acquisition methods, such as fair active learning approaches, typically require annotating sensitive attributes. We propose a tractable active data sampling algorithm that does not rely on training group annotations.
  arXiv  Detail & Related papers  (2024-02-20T07:57:38Z)
• Sample Complexity of Preference-Based Nonparametric Off-Policy Evaluation with Deep Networks [58.469818546042696]
  We study the sample efficiency of OPE with human preference and establish a statistical guarantee for it. By appropriately selecting the size of a ReLU network, we show that one can leverage any low-dimensional manifold structure in the Markov decision process.
  arXiv  Detail & Related papers  (2023-10-16T16:27:06Z)
• Understanding and Mitigating Spurious Correlations in Text Classification with Neighborhood Analysis [69.07674653828565]
  Machine learning models have a tendency to leverage spurious correlations that exist in the training set but may not hold true in general circumstances. In this paper, we examine the implications of spurious correlations through a novel perspective called neighborhood analysis. We propose a family of regularization methods, NFL (doN't Forget your Language) to mitigate spurious correlations in text classification.
  arXiv  Detail & Related papers  (2023-05-23T03:55:50Z)
• Offline RL with No OOD Actions: In-Sample Learning via Implicit Value Regularization [90.9780151608281]
  In-sample learning (IQL) improves the policy by quantile regression using only data samples. We make a key finding that the in-sample learning paradigm arises under the textitImplicit Value Regularization (IVR) framework. We propose two practical algorithms, Sparse $Q$-learning (EQL) and Exponential $Q$-learning (EQL), which adopt the same value regularization used in existing works.
  arXiv  Detail & Related papers  (2023-03-28T08:30:01Z)
• Adaptive Negative Evidential Deep Learning for Open-set Semi-supervised Learning [69.81438976273866]
  Open-set semi-supervised learning (Open-set SSL) considers a more practical scenario, where unlabeled data and test data contain new categories (outliers) not observed in labeled data (inliers) We introduce evidential deep learning (EDL) as an outlier detector to quantify different types of uncertainty, and design different uncertainty metrics for self-training and inference. We propose a novel adaptive negative optimization strategy, making EDL more tailored to the unlabeled dataset containing both inliers and outliers.
  arXiv  Detail & Related papers  (2023-03-21T09:07:15Z)
• Semi-supervised Batch Learning From Logged Data [24.826544828460158]
  We build on the counterfactual risk minimization framework, which also assumes access to propensity scores. We propose learning methods for problems where feedback is missing for some samples, so there are samples with feedback and samples missing-feedback in the logged data.
  arXiv  Detail & Related papers  (2022-09-15T08:58:28Z)
• Out-of-Scope Intent Detection with Self-Supervision and Discriminative Training [20.242645823965145]
  Out-of-scope intent detection is of practical importance in task-oriented dialogue systems. We propose a method to train an out-of-scope intent classifier in a fully end-to-end manner by simulating the test scenario in training. We evaluate our method extensively on four benchmark dialogue datasets and observe significant improvements over state-of-the-art approaches.
  arXiv  Detail & Related papers  (2021-06-16T08:17:18Z)
• DisCor: Corrective Feedback in Reinforcement Learning via Distribution Correction [96.90215318875859]
  We show that bootstrapping-based Q-learning algorithms do not necessarily benefit from corrective feedback. We propose a new algorithm, DisCor, which computes an approximation to this optimal distribution and uses it to re-weight the transitions used for training.
  arXiv  Detail & Related papers  (2020-03-16T16:18:52Z)
• Learning with Out-of-Distribution Data for Audio Classification [60.48251022280506]
  We show that detecting and relabelling certain OOD instances, rather than discarding them, can have a positive effect on learning. The proposed method is shown to improve the performance of convolutional neural networks by a significant margin.
  arXiv  Detail & Related papers  (2020-02-11T21:08:06Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.