Code for the paper "Implicit Representations of Meaning in Neural Language Models"

Last update: Nov 03, 2022

Related tags

Overview

Implicit Representations of Meaning in Neural Language Models

Preliminaries

Create and set up a conda environment as follows:

conda create -n state-probes python=3.7
conda activate state-probes
pip install -r requirements.txt

Install the appropriate torch 1.7.0 for your cuda version:

conda install pytorch==1.7.0 cudatoolkit=<cuda_version> -c pytorch

Before running any command below, run

export PYTHONPATH=.
export TOKENIZERS_PARALLELISM=true

Data

The Alchemy data is downloaded from their website.

wget https://nlp.stanford.edu/projects/scone/scone.zip
unzip scone.zip

The synthetic version of alchemy was generated by running:

echo 0 > id #the code requires a file called id with a number in it ...
python alchemy_artificial_generator.py --num_scenarios 3600 --output synth_alchemy_train
python alchemy_artificial_generator.py --num_scenarios 500 --output synth_alchemy_dev
python alchemy_artificial_generator.py --num_scenarios 900 --output synth_alchemy_test

You can also just download our generated data through:

wget http://web.mit.edu/bzl/www/synth_alchemy.tar.gz
tar -xzvf synth_alchemy.tar.gz

The Textworld data is under

wget http://web.mit.edu/bzl/www/tw_data.tar.gz
tar -xzvf tw_data.tar.gz

LM Training

To train a BART or T5 model on Alchemy data

python scripts/train_alchemy.py \
    --arch [t5|bart] [--no_pretrain] \
    [--synthetic] --encode_init_state NL

Saves model checkpoints under sconeModels/*.

To train a BART or T5 model on Textworld data

python scripts/train_textworld.py \
    --arch [t5/bart] [--no_pretrain] \
    --data tw_data/simple_traces --gamefile tw_games/simple_games

Saves model checkpoints nder twModels/*.

Probe Training & Evaluation

Alchemy

The main probe command is as follows:

python scripts/probe_alchemy.py \
    --arch [bart|t5] --lm_save_path <path_to_lm_checkpoint> [--no_pretrain] \
    --encode_init_state NL --nonsynthetic \
    --probe_target single_beaker_final.NL --localizer_type single_beaker_init_full \
    --probe_type linear --probe_agg_method avg \
    --encode_tgt_state NL.[bart|t5] --tgt_agg_method avg \
    --batchsize 128 --eval_batchsize 1024 --lr 1e-4

For evaluation, add --eval_only --probe_save_path <path_to_probe_checkpoint>. This will save model predictions to a .jsonl file under the same directory as the probe checkpoint.

Add --control_input for No LM experiment.

Change --probe_target to single_beaker_init.NL to decode initial state.

For localization experiments, set --localizer_type single_beaker_init_{$i}.offset{$off} for some token i in {article, pos.[R0|R1|R2], beaker.[R0|R1], verb, amount, color, end_punct} and some integer offset off between 0 and 6.

Saves probe checkpoints under probe_models_alchemy/*.

Intervention experiment results follow from running the script:

python scripts/intervention.py \
    --arch [bart|t5] \
    --encode_init_state NL \
    --create_type drain_1 \
    --lm_save_path <path_to_lm_checkpoint>

which creates two contexts and replaces a select few encoded tokens to modify the underlying belief state.

Textworld

Begin by creating the full set of encoded proposition representations

python scripts/get_all_tw_facts.py \
    --data tw_data/simple_traces --gamefile tw_data/simple_games \
    --state_model_arch [bart|t5] \
    --probe_target belief_facts_pair \
    --state_model_path [None|pretrain|<path_to_lm_checkpoint>] \
    --out_file <path_to_prop_encodings>

Run the probe with

python scripts/probe_textworld.py \
    --arch [bart|t5] --data tw_data/simple_traces --gamefile tw_data/simple_games \
    --probe_target final.full_belief_facts_pair --encode_tgt_state NL.[bart|t5] \
    --localizer_type belief_facts_pair_[first|last|all] --probe_type 3linear_classify \
    --probe_agg_method avg --tgt_agg_method avg \
    --lm_save_path <path_to_lm_checkpoint> [--no_pretrain] \
    --ents_to_states_file <path_to_prop_encodings> \
    --eval_batchsize 256 --batchsize 32

For evaluation, add --eval_only --probe_save_path <path_to_probe_checkpoint>. This will save model predictions to a .jsonl file under the same directory as the probe checkpoint.

Add --control_input for No LM experiment.

Change --probe_target to init.full_belief_facts_pair to decode initial state.

For remap experiments, change --probe_target to final.full_belief_facts_pair.control_with_rooms.

For decoding from just one side of propositions, replace any instance of belief_facts_pair (in --probe_target and --localizer_type) with belief_facts_single and rerun both commands (first get the full proposition encodings, then run the probe).

Saves probe checkpoints under probe_models_textworld/*.

Print Metrics

Print full metrics (state EM, entity EM, subdivided by relations vs. propositions, etc.) using scripts/print_metrics.py.

python scripts/print_metrics.py \
    --arch [bart|t5] --domain [alchemy|textworld] \
    --pred_files <path_to_model_predictions_1>,<path_to_model_predictions_2>,<path_to_model_predictions_3>,... \
    [--use_remap_domain --remap_fn <path_to_remap_model_predictions>] \
    [--single_side_probe]

Code for the paper "Implicit Representations of Meaning in Neural Language Models"

Related tags

Overview

Implicit Representations of Meaning in Neural Language Models

Preliminaries

Data

LM Training

Probe Training & Evaluation

Alchemy

Textworld

Print Metrics

Owner

Belinda Li

i-SpaSP: Structured Neural Pruning via Sparse Signal Recovery

Libraries, tools and tasks created and used at DeepMind Robotics.

AirPose: Multi-View Fusion Network for Aerial 3D Human Pose and Shape Estimation

Code, final versions, and information on the Sparkfun Graphical Datasheets

Asymmetric Bilateral Motion Estimation for Video Frame Interpolation, ICCV2021

Kroomsa: A search engine for the curious

Principled Detection of Out-of-Distribution Examples in Neural Networks

UnivNet: A Neural Vocoder with Multi-Resolution Spectrogram Discriminators for High-Fidelity Waveform Generation

Tool for working with Y-chromosome data from YFull and FTDNA

Locally Differentially Private Distributed Deep Learning via Knowledge Distillation (LDP-DL)

[LREC] MMChat: Multi-Modal Chat Dataset on Social Media

HPRNet: Hierarchical Point Regression for Whole-Body Human Pose Estimation

COVINS -- A Framework for Collaborative Visual-Inertial SLAM and Multi-Agent 3D Mapping

Election Exit Poll Prediction and U.S.A Presidential Speech Analysis using Machine Learning

Unrestricted Facial Geometry Reconstruction Using Image-to-Image Translation

🛠️ Tools for Transformers compression using Lightning ⚡

Put blind watermark into a text with python

Pytorch version of VidLanKD: Improving Language Understanding viaVideo-Distilled Knowledge Transfer

Rank1 Conversation Emotion Detection Task

Council-GAN - Implementation for our paper Breaking the Cycle - Colleagues are all you need (CVPR 2020)