# Data Feedback Loops: Model-driven Amplification of Dataset Biases

Source: [Data Feedback Loops: Model-driven Amplification of Dataset Biases](<https://www.sota2.com/research/paper/68fa6d6a1e209131cea39b09>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-26T00:33:23.458Z

Authors: Rohan Taori, Tatsunori B. Hashimoto

Published: 2022-09-08T17:35:51.000Z

## Abstract

Datasets scraped from the internet have been critical to the successes of large-scale machine learning. Yet, this very success puts the utility of future internet-derived datasets at potential risk, as model outputs begin to replace human annotations as a source of supervision.   In this work, we first formalize a system where interactions with one model are recorded as history and scraped as training data in the future. We then analyze its stability over time by tracking changes to a test-time bias statistic (e.g. gender bias of model predictions). We find that the degree of bias amplification is closely linked to whether the model's outputs behave like samples from the training distribution, a behavior which we characterize and define as consistent calibration. Experiments in three conditional prediction scenarios - image classification, visual role-labeling, and language generation - demonstrate that models that exhibit a sampling-like behavior are more calibrated and thus more stable. Based on this insight, we propose an intervention to help calibrate and stabilize unstable feedback systems.   Code is available at https://github.com/rtaori/data\_feedback.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2209.03942>)

[PDF](<https://arxiv.org/pdf/2209.03942.pdf>)
