pramodkaushik/np_analysis
Adversarial attacks generated for the ACL paper "Did the Model Understand the Question?"
This tool helps AI researchers and developers evaluate how well their neural network models understand questions, particularly in natural language processing tasks. It takes a pre-trained Neural Programmer model and a dataset of questions, then applies adversarial attacks to generate perturbed versions of those questions. The output helps you assess your model's robustness and true understanding of the input, rather than just its performance on original data.
No commits in the last 6 months.
Use this if you are an AI researcher or machine learning engineer studying the interpretability and robustness of neural networks, specifically those designed for question answering on structured data like tables.
Not ideal if you are looking for a general-purpose natural language processing tool for end-user applications or a library for building new neural network architectures from scratch.
Stars
8
Forks
2
Language
Jupyter Notebook
License
Apache-2.0
Category
Last pushed
Jun 23, 2018
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/nlp/pramodkaushik/np_analysis"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
thunlp/OpenAttack
An Open-Source Package for Textual Adversarial Attack.
thunlp/TAADpapers
Must-read Papers on Textual Adversarial Attack and Defense
jind11/TextFooler
A Model for Natural Language Attack on Text Classification and Inference
thunlp/OpenBackdoor
An open-source toolkit for textual backdoor attack and defense (NeurIPS 2022 D&B, Spotlight)
thunlp/HiddenKiller
Code and data of the ACL-IJCNLP 2021 paper "Hidden Killer: Invisible Textual Backdoor Attacks...