Understanding Programmatic Weak Supervision via Source-aware Influence Function
This provides a tool for users of weak supervision to debug and improve their pipelines, though it is incremental as it builds on existing Influence Function methods.
The paper tackled the problem of interpreting and debugging Programmatic Weak Supervision (PWS) pipelines by proposing source-aware Influence Function, which decomposes influences of components like source votes and training data. The result demonstrated use cases including identifying mislabeling with 9%-37% gains over baselines and improving model generalization by 13%-24% over ordinary IF.
Programmatic Weak Supervision (PWS) aggregates the source votes of multiple weak supervision sources into probabilistic training labels, which are in turn used to train an end model. With its increasing popularity, it is critical to have some tool for users to understand the influence of each component (e.g., the source vote or training data) in the pipeline and interpret the end model behavior. To achieve this, we build on Influence Function (IF) and propose source-aware IF, which leverages the generation process of the probabilistic labels to decompose the end model's training objective and then calculate the influence associated with each (data, source, class) tuple. These primitive influence score can then be used to estimate the influence of individual component of PWS, such as source vote, supervision source, and training data. On datasets of diverse domains, we demonstrate multiple use cases: (1) interpreting incorrect predictions from multiple angles that reveals insights for debugging the PWS pipeline, (2) identifying mislabeling of sources with a gain of 9%-37% over baselines, and (3) improving the end model's generalization performance by removing harmful components in the training objective (13%-24% better than ordinary IF).