paper-with-me

홈 › Papers

Hidden Incentives for Auto-Induced Distributional Shift

2020-09-19 · David Krueger, Tegan Maharaj, Jan Leike

Decisions made by machine learning systems have increasing influence on the world, yet it is common for machine learning algorithms to assume that no such influence exists. An example is the use of the i.i.d. assumption in content recommendation. In fact, the (choice of) content displayed can change users' perceptions and preferences, or even drive them away, causing a shift in the distribution of users. We introduce the term auto-induced distributional shift (ADS) to describe the phenomenon of an algorithm causing a change in the distribution of its own inputs. Our goal is to ensure that machine learning systems do not leverage ADS to increase performance when doing so could be undesirable. We demonstrate that changes to the learning algorithm, such as the introduction of meta-learning, can cause hidden incentives for auto-induced distributional shift (HI-ADS) to be revealed. To address this issue, we introduce `unit tests' and a mitigation strategy for HI-ADS, as well as a toy environment for modelling real-world issues with HI-ADS in content recommendation, where we demonstrate that strong meta-learners achieve gains in performance via ADS. We show meta-learning and Q-learning both sometimes fail unit tests, but pass when using our mitigation strategy.

📄 PDF Abstract BibTeX arXiv:2009.09153

Code (0)

등록된 구현이 없습니다.

Tasks

BIG-bench Machine LearningMeta-LearningQ-Learning

Methods 이 논문이 사용한 방법론

Q-Learning Q-Learning is an off-policy temporal difference control algorithm: $$Q\left(S\_{t}, A\_{t}\right) \leftarrow Q\left(S\_{t}, A\_{t}\right) + \alpha\left[R_{t+1} +…

Similar Papers 제목 키워드 기반

Hidden incentives for self-induced distributional shift

2019-09-25 · David Scott Krueger, Tegan Maharaj, Shane Legg, Jan Leike

Decisions made by machine learning systems have increasing influence on the world. Yet it is common for machine learning algorithms to assume that no such influence exists. An example is the use of the i.i.d. assumption …

BIG-bench Machine LearningMeta-Learning

Revealing the Incentive to Cause Distributional Shift

2021-09-29 · David Krueger, Tegan Maharaj, Jan Leike

Decisions made by machine learning systems have increasing influence on the world, yet it is common for machine learning algorithms to assume that no such influence exists. An example is the use of the i.i.d. assumption …

Meta-Learning

Distributionally Robust Imitation Learning: Layered Control Architecture for Certifiable Autonomy

2025-12-19 · Aditya Gahlawat, Ahmed Aboudonia, Sandeep Banik, Naira Hovakimyan 외 arxiv

Imitation learning (IL) enables autonomous behavior by learning from expert demonstrations. While more sample-efficient than comparative alternatives like reinforcement learning, IL is sensitive to compounding errors ind…

Reinforcement Learning

The Hidden Puppet Master: Predicting Human Belief Change in Manipulative LLM Dialogues

2026-03-21 · Jocelyn Shen, Amina Luvsanchultem, Jessica Kim, Kynnedy Smith 외 arxiv

As users increasingly turn to LLMs for practical and personal advice, they become vulnerable to subtle steering toward hidden incentives misaligned with their own interests. While existing NLP research has benchmarked ma…

Distributional Shift-Aware Off-Policy Interval Estimation: A Unified Error Quantification Framework

2023-09-23 · Wenzhuo Zhou, Yuhan Li, Ruoqing Zhu, Annie Qu

We study high-confidence off-policy evaluation in the context of infinite-horizon Markov decision processes, where the objective is to establish a confidence interval (CI) for the target policy value using only offline d…

Off-policy evaluation