Probing Neural Network Understanding of Natural Language Arguments

Link

Authors: Timothy Niven and Hung-Yu Kao

Abstract:

We are surprised to find that BERT's peak performance of 77% on the Argument Reasoning Comprehension Task reaches just three points below the average untrained human baseline. However, we show that this result is entirely accounted for by exploitation of spurious statistical cues in the dataset. We analyze the nature of these cues and demonstrate that a range of models all exploit them. This analysis informs the construction of an adversarial dataset on which all models achieve random accuracy. Our adversarial dataset provides a more robust assessment of argument comprehension and should be adopted as the standard in future work.

Reference:

@inproceedings{niven-kao-2019-probing,
    title = "Probing Neural Network Comprehension of Natural Language Arguments",
    author = "Niven, Timothy  and
      Kao, Hung-Yu",
    booktitle = "Proceedings of the 57th Conference of the Association for Computational Linguistics",
    month = jul,
    year = "2019",
    address = "Florence, Italy",
    publisher = "Association for Computational Linguistics",
    url = "https://www.aclweb.org/anthology/P19-1459",
    pages = "4658--4664",
    abstract = "We are surprised to find that BERT{'}s peak performance of 77{\%} on the Argument Reasoning Comprehension Task reaches just three points below the average untrained human baseline. However, we show that this result is entirely accounted for by exploitation of spurious statistical cues in the dataset. We analyze the nature of these cues and demonstrate that a range of models all exploit them. This analysis informs the construction of an adversarial dataset on which all models achieve random accuracy. Our adversarial dataset provides a more robust assessment of argument comprehension and should be adopted as the standard in future work.",
}

Errata

Due to an errata in the original paper, we have updated the arXiv version here. The errata was that our adversarial experiments used the original (not claim-negated) data for training, augmented with swapped warrants. This is important for breaking the validity of spurious cues over the claims and warrants. The negated claim augmentation has successfully eliminated warrant cues as a solution. Details of the problem of additional cues over claims and warrants will be provided in a forthcoming paper.

Adversarial Dataset

Provided in the adversarial_dataset folder. The setup we used in the experiments is

• train-swapped.csv
• dev-adv-negated.csv
• test-adv-negated.csv

Viewing our Results

Each experiment has its own folder in the results_from_paper folder. The suffixes (and combinations thereof) indicate the setup

• cw only considers claims and warrants
• rw only considers reasons and warrants
• w only considers warrants
• adv uses the adversarial dev and test dataset, and swap augmented train dataset The tables below also show the experiment names for reproducing specific results.

Within each experiment's folder in results you will find

• accs.csv: contains accuracies for train, dev, and test over all random seeds
• best_params.json: lists the best parameters from grid search
• grid.csv: lists all grid search results and parameter combinations
• preds.csv: lists all predictions for all data points which can be filtered by dataset and queried by each data point's unique identifier

You can get a summary of the accuracies over various random seeds for an experiment by running

python accs.py experiment_name --from_paper

For details of how each experiment is run, you can view the files in the experiments folder.

The from_paper argument will pull results from the results_from_paper directory. Without this flag it will read the results directory, where any experiment results you run yourself locally will be recorded.

Reproducing our Results

Virtual Environment

Package requirements are listed in requirements.txt. We used and tested this repository with Python 3.6. To simplify this, here are the
commands I issued on my Ubuntu computer to make this repository work:

• conda create --name arct2 python=3.6
• conda activate arct2
• pip install pandas==0.23.4
• pip install nltk==3.4
• pip install tqdm==4.28.1
• conda install -c pytorch pytorch=0.4.1
• pip install numpy==1.15.4 (not sure if this is necessary, but it does work)
• pip install pytorch-pretrained-bert==0.1.2

Download and Prepare Data

Run prepare.sh. If you download a new version of this repository, you should run this again.

Running Experiments

Then to reproduce the results of any of the experiments run the script

python run.py experiment_name

Note that due to a bug in the random seed control in the original experiments the exact numbers from the paper cannot be reproduced (the data loaders were initialized before the seed was set). Having fixed the bug we are reproducing the experiments and reporting the exact values you should see when you run this code. As expected they are slightly different, but not qualitatively so. Apparently the order in which the examples are presented does have some effect. BERT Base has so far been able to get a lucky run up to 72.5% on the original dataset.

Results

Table 1

Model	Experiment Name	Dev (Mean)	Test (Mean)	Test (Median)	Test (Max)
Human (trained)			0.909 +/- 0.11
Human (untrained)			0.798 +/- 0.16
BERT (Large)	bert_large	0.694 +/- 0.04	0.660 +/- 0.08	0.690	0.775
GIST (Choi and Lee, 2018)		0.716 +/- 0.01	0.711 +/- 0.01
BERT (Base)	bert_base	0.675 +/- 0.03	0.634 +/- 0.07	0.661	0.725
World Knowledge (Botschen et al., 2018)		0.674 +/- 0.01	0.568 +/- 0.03		0.610
Bag of Vectors (BoV)	bov	0.633 +/- 0.02	0.564 +/- 0.02	0.562	0.604
BiLSTM	bilstm	0.659 +/- 0.01	0.544 +/- 0.02	0.547	0.583

Table 4

Model	Experiment Name	Adversarial Test (Mean)	Adversarial Test (Median)	Adversarial Test (Max)
BERT (Large)	bert_large_adv	0.509 +/- 0.02	0.509	0.539
BoV	bov_adv	0.500 +/- 0.00	0.500	0.503
BiLSTM	bilstm_adv	0.499 +/- 0.00	0.500	0.501