# WheaCha: A Method for Explaining the Predictions of Models of Code

arXiv: https://arxiv.org/abs/2102.04625 · HTML: https://mlsanity.com/paper/2102.04625/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 21 of its 44 references that are in the map.
- First posted on arXiv 9 Feb 2021.

## Builds on (most cited first, 21 of 21)

- [A Unified Approach to Interpreting Model Predictions](https://mlsanity.com/paper/1705.07874/) (2017)
- ["Why Should I Trust You?": Explaining the Predictions of Any Classifier](https://mlsanity.com/paper/1602.04938/) (2016)
- [Neural Machine Translation by Jointly Learning to Align and Translate](https://mlsanity.com/paper/1409.0473/) (2014)
- [Axiomatic Attribution for Deep Networks](https://mlsanity.com/paper/1703.01365/) (2017)
- [Deep Inside Convolutional Networks: Visualising Image Classification Models and Saliency Maps](https://mlsanity.com/paper/1312.6034/) (2013)
- [Learning Important Features Through Propagating Activation Differences](https://mlsanity.com/paper/1704.02685/) (2017)
- [CodeBERT: A Pre-Trained Model for Programming and Natural Languages](https://mlsanity.com/paper/2002.08155/) (2020)
- [CodeSearchNet Challenge: Evaluating the State of Semantic Code Search](https://mlsanity.com/paper/1909.09436/) (2019)
- [A Benchmark for Interpretability Methods in Deep Neural Networks](https://mlsanity.com/paper/1806.10758/) (2018)
- [Not Just a Black Box: Learning Important Features Through Propagating Activation Differences](https://mlsanity.com/paper/1605.01713/) (2016)
- [code2vec: Learning Distributed Representations of Code](https://mlsanity.com/paper/1803.09473/) (2018)
- [Learning to Represent Programs with Graphs](https://mlsanity.com/paper/1711.00740/) (2017)
- [code2seq: Generating Sequences from Structured Representations of Code](https://mlsanity.com/paper/1808.01400/) (2018)
- [Visualizing Deep Neural Network Decisions: Prediction Difference Analysis](https://mlsanity.com/paper/1702.04595/) (2017)
- [SequenceR: Sequence-to-Sequence Learning for End-to-End Program Repair](https://mlsanity.com/paper/1901.01808/) (2018)
- [Structured Neural Summarization](https://mlsanity.com/paper/1811.01824/) (2018)
- [LambdaNet: Probabilistic Type Inference using Graph Neural Networks](https://mlsanity.com/paper/2005.02161/) (2020)
- [On the Generalizability of Neural Program Models with respect to Semantic-Preserving Program Transformations](https://mlsanity.com/paper/2008.01566/) (2020)
- [Dynamic Neural Program Embedding for Program Repair](https://mlsanity.com/paper/1711.07163/) (2017)
- [Learning Semantic Program Embeddings with Graph Interval Neural Network](https://mlsanity.com/paper/2005.09997/) (2020)
- [Learning Nonlinear Loop Invariants with Gated Continuous Logic Networks (Extended Version)](https://mlsanity.com/paper/2003.07959/) (2020)

## Built on by (most cited first, 1 of 1)

- [Study of Distractors in Neural Models of Code](https://mlsanity.com/paper/2303.01739/) (2023)

## Abstract (arXiv)

Attribution methods have emerged as a popular approach to interpreting model predictions based on the relevance of input features. Although the feature importance ranking can provide insights of how models arrive at a prediction from a raw input, they do not give a clear-cut definition of the key features models use for the prediction. In this paper, we present a new method, called WheaCha, for explaining the predictions of code models. Although WheaCha employs the same mechanism of tracing model predictions back to the input features, it differs from all existing attribution methods in crucial ways. Specifically, WheaCha divides an input program into "wheat" (i.e., the defining features that are the reason for which models predict the label that they predict) and the rest "chaff" for any prediction of a learned code model. We realize WheaCha in a tool, HuoYan, and use it to explain four prominent code models: code2vec, seq-GNN, GGNN, and CodeBERT. Results show (1) HuoYan is efficient - taking on average under twenty seconds to compute the wheat for an input program in an end-to-end fashion (i.e., including model prediction time); (2) the wheat that all models use to predict input programs is made of simple syntactic or even lexical properties (i.e., identifier names); (3) Based on wheat, we present a novel approach to explaining the predictions of code models through the lens of training data.