Codebase for testing whether hidden states of neural networks encode discrete structures.

Overview

A vector space and a parse tree representing the same info.

structural-probes

Codebase for testing whether hidden states of neural networks encode discrete structures.

Based on the paper A Structural Probe for Finding Syntax in Word Representations.

See the blog post on structural probes for a brief introduction.

Installing & Getting Started

  1. Clone the repository.

     git clone https://github.com/john-hewitt/structural-probes/
     cd structural-probes
    
  2. [Optional] Construct a virtual environment for this project. Only python3 is supported.

     conda create --name sp-env
     conda activate sp-env
    
  3. Install the required packages. This mainly means pytorch, scipy, numpy, seaborn, etc. Look at pytorch.org for the PyTorch installation that suits you and install it; it won't be installed via requirements.txt. Everything in the repository will use a GPU if available, but if none is available, it will detect so and just use the CPU, so use the pytorch install of your choice.

     conda install --file requirements.txt
     pip install pytorch-pretrained-bert
    
  4. Download some pre-packaged data from the English Universal Dependencies (EWT) dataset and pretrained probes to get your feet wet.

     bash ./download_example.sh
    

    This will make the directory example/data, and in it will be 9 files, 3 for each of train,dev,test.

    • en_ewt-ud-{train,dev,test}.conllu: the parsed language data
    • en_ewt-ud-{train,dev,test}.txt: whitespace-tokenized, sentence-per-line language data.
    • en_ewt-ud-{train,dev,test}.elmo-layers.hdf5: the ELMo hidden states for each sentence of the language data, constructed by running elmo on the .txt files.
  5. Test a pre-trained structural probe on BERTlarge with our demo script!

     printf "The chef that went to the stores was out of food" | python structural-probes/run_demo.py example/demo-bert.yaml
    

    The script will make a new directory under example/results/ and store some neat visualizations there. It will use pre-trained probe parameters stored at example/data, downloaded with download_example.sh. Try out some other sentences too!

  6. Run an experiment using an example experiment configuration, and take a look at the resultant reporting!

     python structural-probes/run_experiment.py example/config/prd_en_ewt-ud-sample.yaml
    

    The path to a new directory containing the results of the experiment will be in the first few lines of the logging output of the script. Once you go there, you can see dev-pred*.png: some distance matrices printed by the script, as well as files containing the quantitative reporting results, like dev.uuas, the unlabeled undirected attachment score. These will all be very low, since the probe was trained on very little data!

Run a pretrained structural probe on BERT-large quickly on the command line.

It's easy to get predictions on a sentence (or file of sentences) using our demo script and the pre-trained structural probes we release. We use pytorch-pretrained-bert to get BERT subword embeddings for each sentence; it should be installed during setup of the repository.

Make sure you've run download_example.sh; this will download two probe parameter files to example/data/. Also make sure you've installed all dependencies. One is a distance probe on the 16th hidden layer of BERT large, and the other is a depth probe on the same layer. The configuration file example/demo-bert.yaml has the right paths already plugged in; just pipe text into the demo file, as follows:

 printf "The chef that went to the stores was out of food" | python structural-probes/run_demo.py example/demo-bert.yaml

If you want to run multiple sentences at once, you can either do so via printf:

 printf "The chef that went to the stores was out of food\nThe chef that went to the stores and talked to the parents was out of food" | python structural-probes/run_demo.py example/demo-bert.yaml

Or piping/redirecting a file to stdin:

 cat my_set.txt | python structural-probes/run_demo.py example/demo-bert.yaml

The script will print out a directory to which it has written visualizations of both parse depths and parse distances as predicted by a distance probe and a depth probe. You'll also see demo.tikz, which is a bit of LaTeX for the tikz-dependency package. With tikz-dependency in the same directory as your LaTeX file, you can plop this bit of LaTeX in a figure environment and see the minimum spanning tree it constructs. It'd look a bit like this:

\documentclass{article}
\usepackage{tikz-dependency}
\usepackage{tikz}

\pgfkeys{%
/depgraph/reserved/edge style/.style = {% 
white, -, >=stealth, % arrow properties                                                                            
black, solid, line cap=round, % line properties
rounded corners=2, % make corners round
},%
}
\begin{document}
\begin{figure}
  \centering
  \small
  \begin{dependency}[hide label, edge unit distance=.5ex]
    \begin{deptext}[column sep=0.05cm]
      The\& chef\& who\& ran\& to\& the\& stores\& is\& out\& of\& food \\
    \end{deptext}                                                                                                                                                                                                                           
    \depedge[edge style={red}, edge below]{8}{9}{.}
    \depedge[edge style={red}, edge below]{5}{7}{.}
    \depedge[edge style={red}, edge below]{4}{5}{.}
    \depedge[edge style={red}, edge below]{1}{2}{.}
    \depedge[edge style={red}, edge below]{6}{7}{.}
    \depedge[edge style={red}, edge below]{9}{10}{.}
    \depedge[edge style={red}, edge below]{10}{11}{.}
    \depedge[edge style={red}, edge below]{3}{4}{.}
    \depedge[edge style={red}, edge below]{2}{4}{.}
    \depedge[edge style={red}, edge below]{2}{8}{.}
  \end{dependency}
\end{figure}
\end{document}

Which results in a PDF with the following:

A predicted parse tree for the sentence 'The chef who ran to the stores is out of food'

Note that your text should be whitespace-tokenized! If you want to evaluate on a test set with gold parses, or if you want to train your own structural probes, read on.

The experiment config file

Experiments run with this repository are specified via yaml files that completely describe the experiment (except the random seed.) In this section, we go over each top-level key of the experiment config.

Dataset:

  • observation_fieldnames: the fields (columns) of the conll-formatted corpus files to be used. Must be in the same order as the columns of the corpus. Each field will be accessable as an attribute of each Observation class (e.g., observation.sentence contains the sequence of tokens comprising the sentence.)
  • corpus: The location of the train, dev, and test conll-formatted corpora files. Each of train_path, dev_path, test_path will be taken as relative to the root field.
  • embeddings: The location of the train, dev, and test pre-computed embedding files (ignored if not applicable. Each of train_path, dev_path, test_path will be taken as relative to the root field. - type is ignored.
  • batch_size: The number of observations to put into each batch for training the probe. 20 or so should be great.
dataset:
  observation_fieldnames:
     - index
     - sentence
     - lemma_sentence
     - upos_sentence
     - xpos_sentence
     - morph
     - head_indices
     - governance_relations
     - secondary_relations
     - extra_info
     - embeddings
  corpus:
    root: example/data/en_ewt-ud-sample/
    train_path: en_ewt-ud-train.conllu
    dev_path: en_ewt-ud-dev.conllu
    test_path: en_ewt-ud-test.conllu
  embeddings:
    type: token #{token,subword}
    root: example/data/en_ewt-ud-sample/ 
    train_path: en_ewt-ud-train.elmo-layers.hdf5
    dev_path: en_ewt-ud-dev.elmo-layers.hdf5
    test_path: en_ewt-ud-test.elmo-layers.hdf5
  batch_size: 40

Model

  • hidden_dim: The dimensionality of the representations to be probed. The probe parameters constructed will be of shape (hidden_dim, maximum_rank)
  • embedding_dim: ignored
  • model_type: One of ELMo-disk, BERT-disk, ELMo-decay, ELMo-random-projection as of now. Used to help determine which Dataset class should be constructed, as well as which model will construct the representations for the probe. The Decay0 and Proj0 baselines in the paper are from ELMo-decay and ELMo-random-projection, respectively. In the future, will be used to specify other PyTorch models.
  • use_disk: Set to True to assume that pre-computed embeddings should be stored with each Observation; Set to False to use the words in some downstream model (this is not supported yet...)
  • model_layer: The index of the hidden layer to be used by the probe. For example, ELMo models can use layers 0,1,2; BERT-base models have layers 0 through 11; BERT-large 0 through 23.
  • tokenizer: If a model will be used to construct representations on the fly (as opposed to using embeddings saved to disk) then a tokenizer will be needed. The type string will specify the kind of tokenizer used. The vocab_path is the absolute path to a vocabulary file to be used by the tokenizer.
model:
  hidden_dim: 1024 # ELMo hidden dim
  #embedding_dim: 1024 # ELMo word embedding dim
  model_type: ELMo-disk # BERT-disk, ELMo-disk,
  tokenizer:
    type: word
    vocab_path: example/vocab.vocab
  use_disk: True
  model_layer: 2 # BERT-base: {1,...,12}; ELMo: {1,2,3}

Probe, probe-training

  • task_signature: Specifies the function signature of the task. Currently, can be either word, for parse depth (or perhaps labeling) tasks; or word_pair for parse distance tasks.
  • task_name: A unique name for each task supported by the repository. Right now, this includes parse-depth and parse-distance.
  • maximum_rank: Specifies the dimensionality of the space to be projected into, if psd_parameters=True. The projection matrix is of shape (hidden_dim, maximum_rank). The rank of the subspace is upper-bounded by this value. If psd_parameters=False, then this is ignored.
  • psd_parameters: though not reported in the paper, the parse_distance and parse_depth tasks can be accomplished with a non-PSD matrix inside the quadratic form. All experiments for the paper were run with psd_parameters=True, but setting psd_parameters=False will simply construct a square parameter matrix. See the docstring of probe.TwoWordNonPSDProbe and probe.OneWordNonPSDProbe for more info.
  • diagonal: Ignored.
  • prams_path: The path, relative to args['reporting']['root'], to which to save the probe parameters.
  • epochs: The maximum number of epochs to which to train the probe. (Regardless, early stopping is performed on the development loss.)
  • loss: A string to specify the loss class. Right now, only L1 is available. The class within loss.py will be specified by a combination of this and the task name, since for example distances and depths have different special requirements for their loss functions.
probe:
  task_signature: word_pair # word, word_pair
  task_name: parse-distance
  maximum_rank: 32
  psd_parameters: True
  diagonal: False
  params_path: predictor.params
probe_training:
  epochs: 30
  loss: L1

Reporting

  • root: The path to the directory in which a new subdirectory should be constructed for the results of this experiment.
  • observation_paths: The paths, relative to root, to which to write the observations formatted for quick reporting later on.
  • prediction_paths: The paths, relative to root, to which to write the predictions of the model.
  • reporting_methods: A list of strings specifying the methods to use to report and visualize results from the experiment. For parse-distance, the valid methods are spearmanr, uuas, write_predictions, and image_examples. When reporting uuas, some tikz-dependency examples are written to disk as well. For parse-depth, the valid methods are spearmanr, root_acc, write_predictions, and image_examples. Note that image_examples will be ignored for the test set.
reporting:
  root: example/results
  observation_paths:
    train_path: train.observations
    dev_path: dev.observations
    test_path: test.observations
  prediction_paths:
    train_path: train.predictions
    dev_path: dev.predictions
    test_path: test.predictions
  reporting_methods:
    - spearmanr
      #- image_examples
    - uuas

Reporting + visualization

It can be time-consuming to make nice visualizations and make sense of the results from a probing experiment, so this repository does a bit of work for you. This section goes over each of the reporting methods available (under args['reporting']['reporting_methods'] in the experiment config), and exmaples of results.

  • spearmanr: This reporting method calculates the spearman correlation between predicted (distances/depths) and true (distances/depths) as defined by gold parse trees. See the paper or reporting.py docstrings for specifics. With this option enabled, you'll see dev.spearmanr, a TSV with an average Spearman correlation for each sentence length represented in the dev set, as well as dev.spearmanr-5-50-mean, which averages the sentence-average values for all sentence lengths between 5 and 50 (inclusive.)
  • image_examples: This reporting method prints out true and predicted distance matrices as pngs for the first 20 examples in the split. These will be labeled dev-pred0.png, dev-gold0.png, etc. They'll look something like this:

A distance matrix defined by a human-constructed parse tree A distance matrix predicted by a structural probe

  • uuas: This reporting method (only used by parse-distance tasks) will print the unlabeled undirected attachment score to dev.uuas, and write the first 20 development examples' minimum spanning trees (for both gold and predicted distance matrices) in a tikz-dependency LaTeX code format, to dev.tikz. Each sentence can be copy-pasted into a LaTeX doc for visualization. Then they'l look something like this:

    Parse trees constructed as minimum-spanning-trees of gold and predicted distance matrices.

  • root_acc: This reporting method (only used by parse-depth tasks) will print to dev.root_acc the percentage of sentences where the least-deep word in the gold tree (the root) is also the least-deep according to the predicted depths.

Replicating PTB Results for the NAACL'19 Paper

As usual with the PTB, a bit of work has to be done in prepping data (and you have to have the unadulterated PTB data already, not the mangled language modeling benchmark version.)

To replicate our results on the PTB, you'll have to prep some data files. The prep scripts will need to be modified to use paths on your system, but the process is as follows:

  1. Have Stanford CoreNLP installed / on your java classpath, and have allennlp installed.
  2. Convert the PTB constituency trees to Stanford Dependencies in conllx format, using the script scripts/convert_splits_to_depparse.sh. This will write a single conllx file for each of train/dev/test. (This uses CoreNLP.)
  3. Convert the conllx files to sentence-per-line whitespace-tokenized files, using scripts/convert_conll_to_raw.py.
  4. Use scripts/convert_raw_to_bert.py and scripts/convert_raw_to_elmo.sh to take the sentencep-er-line whitespace-tokenized files and write BERT and ELMo vectors to disk in hdf5 format.
  5. Replace the data paths (and choose a results path) in the yaml configs in example/config/naacl19/*/* with the paths that point to your conllx and .hdf5 files as constructed in the above steps. These 118 experiment files specify the configuration of all the experiments that end up in the paper.

Experiments on new datasets or models

In the future I hope to streamline support for plugging in arbitrary PyTorch models in model.py, but because of subword models, tokenization, batching etc. this is beyond my current scope.

Right now, the official way to run experiments on new datasets and representation learners is:

  1. Have a conllx file for the train, dev, and test splits of your dataset.
  2. Write contextual word representations to disk for each of the train, dev, and test split in hdf5 format, where the index of the sentence in the conllx file is the key to the hdf5 dataset object. That is, your dataset file should look a bit like {'0': <np.ndarray(size=(1,SEQLEN1,FEATURE_COUNT))>, '1':<np.ndarray(size=(1,SEQLEN1,FEATURE_COUNT))>...}, etc. Note here that SEQLEN for each sentence must be the number of tokens in the sentence as specified by the conllx file.
  3. Edit a config file from example/config to match the paths to your data, as well as the hidden dimension and labels for the columns in the conllx file. Look at the experiment config section of this README for more information therein. One potential gotcha is that you must have an xpos_sentence field in your conllx (as labeled by your yaml config) since this will be used at evaluation time.

Citation

If you use this repository, please cite:

  @InProceedings{hewitt2019structural,
    author =      "Hewitt, John and Manning, Christopher D.",
    title =       "A Structural Probe for Finding Syntax in Word Representations",
    booktitle =   "North American Chapter of the Association for Computational Linguistics: Human Language Technologies",
    year =        "2019",
    publisher =   "Association for Computational Linguistics",
    location =    "Minneapolis, USA",
  }
Owner
John Hewitt
I'm a PhD student working on: NLP, structure, graphs, bash scripts, RNNs, multilinguality, and teaching others to do the same.
John Hewitt
[NeurIPS 2021] Large Scale Learning on Non-Homophilous Graphs: New Benchmarks and Strong Simple Methods

Large Scale Learning on Non-Homophilous Graphs: New Benchmarks and Strong Simple Methods Large Scale Learning on Non-Homophilous Graphs: New Benchmark

60 Jan 03, 2023
Coarse implement of the paper "A Simultaneous Denoising and Dereverberation Framework with Target Decoupling", On DNS-2020 dataset, the DNSMOS of first stage is 3.42 and second stage is 3.47.

SDDNet Coarse implement of the paper "A Simultaneous Denoising and Dereverberation Framework with Target Decoupling", On DNS-2020 dataset, the DNSMOS

Cyril Lv 43 Nov 21, 2022
Vit-ImageClassification - Pytorch ViT for Image classification on the CIFAR10 dataset

Vit-ImageClassification Introduction This project uses ViT to perform image clas

Kaicheng Yang 4 Jun 01, 2022
Script that receives an Image (original) and a set of images to be used as "pixels" in reconstruction of the Original image using the set of images as "pixels"

picinpics Script that receives an Image (original) and a set of images to be used as "pixels" in reconstruction of the Original image using the set of

RodrigoCMoraes 1 Oct 24, 2021
A universal framework for learning timestamp-level representations of time series

TS2Vec This repository contains the official implementation for the paper Learning Timestamp-Level Representations for Time Series with Hierarchical C

Zhihan Yue 284 Dec 30, 2022
MIRACLE (Missing data Imputation Refinement And Causal LEarning)

MIRACLE (Missing data Imputation Refinement And Causal LEarning) Code Author: Trent Kyono This repository contains the code used for the "MIRACLE: Cau

van_der_Schaar \LAB 15 Dec 29, 2022
Pytorch implementation for Semantic Segmentation/Scene Parsing on MIT ADE20K dataset

Semantic Segmentation on MIT ADE20K dataset in PyTorch This is a PyTorch implementation of semantic segmentation models on MIT ADE20K scene parsing da

MIT CSAIL Computer Vision 4.5k Jan 08, 2023
This respository includes implementations on Manifoldron: Direct Space Partition via Manifold Discovery

Manifoldron: Direct Space Partition via Manifold Discovery This respository includes implementations on Manifoldron: Direct Space Partition via Manifo

dayang_wang 4 Apr 28, 2022
CellRank's reproducibility repository.

CellRank's reproducibility repository We believe that reproducibility is key and have made it as simple as possible to reproduce our results. Please e

Theis Lab 8 Oct 08, 2022
Count GitHub Stars ⭐

Count GitHub Stars per Day ⭐ Track GitHub stars per day over a date range to measure the open-source popularity of different repositories. Requirement

Ultralytics 20 Nov 20, 2022
Betafold - AlphaFold with tunings

BetaFold We (hegelab.org) craeted this standalone AlphaFold (AlphaFold-Multimer,

2 Aug 11, 2022
This is the repository for the paper "Have I done enough planning or should I plan more?"

Metacognitive Learning Tool box https://re.is.mpg.de What Is This? This repository contains two modules used to analyse metacognitive learning in huma

0 Dec 01, 2021
[ICRA2021] Reconstructing Interactive 3D Scene by Panoptic Mapping and CAD Model Alignment

Interactive Scene Reconstruction Project Page | Paper This repository contains the implementation of our ICRA2021 paper Reconstructing Interactive 3D

97 Dec 28, 2022
Semi-supevised Semantic Segmentation with High- and Low-level Consistency

Semi-supevised Semantic Segmentation with High- and Low-level Consistency This Pytorch repository contains the code for our work Semi-supervised Seman

123 Dec 30, 2022
Source Code for our paper: Understand me, if you refer to Aspect Knowledge: Knowledge-aware Gated Recurrent Memory Network

KaGRMN-DSG_ABSA This repository contains the PyTorch source Code for our paper: Understand me, if you refer to Aspect Knowledge: Knowledge-aware Gated

XingBowen 4 May 20, 2022
A list of all papers and resoureces on Semantic Segmentation

Semantic-Segmentation A list of all papers and resoureces on Semantic Segmentation. Dataset importance SemanticSegmentation_DL Some implementation of

Alan Tang 1.1k Dec 12, 2022
JUSTICE: A Benchmark Dataset for Supreme Court’s Judgment Prediction

JUSTICE: A Benchmark Dataset for Supreme Court’s Judgment Prediction CSCI 544 Final Project done by: Mohammed Alsayed, Shaayan Syed, Mohammad Alali, S

Smit Patel 3 Dec 28, 2022
Bidimensional Leaderboards: Generate and Evaluate Language Hand in Hand

Bidimensional Leaderboards: Generate and Evaluate Language Hand in Hand Introduction We propose a generalization of leaderboards, bidimensional leader

4 Dec 03, 2022
COIN the currently largest dataset for comprehensive instruction video analysis.

COIN Dataset COIN is the currently largest dataset for comprehensive instruction video analysis. It contains 11,827 videos of 180 different tasks (i.e

86 Dec 28, 2022
Source code for Zalo AI 2021 submission

zalo_ltr_2021 Source code for Zalo AI 2021 submission Solution: Pipeline We use the pipepline in the picture below: Our pipeline is combination of BM2

128 Dec 27, 2022