Adversarial NLP Robustness NLP Tools

Tools, datasets, and methods for generating adversarial attacks on NLP models, evaluating model robustness, and developing defense mechanisms against adversarial perturbations. Does NOT include general model evaluation, bias detection without adversarial intent, or data augmentation for improving performance.

There are 39 adversarial nlp robustness tools tracked. 1 score above 50 (established tier). The highest-rated is thunlp/OpenAttack at 56/100 with 772 stars and 218 monthly downloads.

Get all 39 projects as JSON

curl "https://pt-edge.onrender.com/api/v1/datasets/quality?domain=nlp&subcategory=adversarial-nlp-robustness&limit=20"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.

#	Tool	Score	Tier	Stars	Language
1	thunlp/OpenAttack An Open-Source Package for Textual Adversarial Attack.	56	Established	772	Python
2	thunlp/TAADpapers Must-read Papers on Textual Adversarial Attack and Defense	42	Emerging	1,574	Python
3	osoleve/glitchlings Enemies for your LLM	42	Emerging	35	Python
4	jind11/TextFooler A Model for Natural Language Attack on Text Classification and Inference	40	Emerging	529	Python
5	thunlp/OpenBackdoor An open-source toolkit for textual backdoor attack and defense (NeurIPS 2022...	36	Emerging	200	Python
6	thunlp/SememePSO-Attack Code and data of the ACL 2020 paper "Word-level Textual Adversarial...	34	Emerging	88	Python
7	thunlp/HiddenKiller Code and data of the ACL-IJCNLP 2021 paper "Hidden Killer: Invisible Textual...	34	Emerging	43	Python
8	aonotas/adversarial_text Code for Adversarial Training Methods for Semi-Supervised Text Classification	31	Emerging	124	Python
9	thunlp/Advbench Code and data of the EMNLP 2022 paper "Why Should Adversarial Perturbations...	31	Emerging	72	Python
10	thunlp/BkdAtk-LWS Code and data of the ACL 2021 paper "Turn the Combination Lock: Learnable...	31	Emerging	16	Python
11	salesforce/adversarial-polyglots Code for the paper "Code-Mixing on Sesame Street: Dawn of the Adversarial...	30	Emerging	10	Python
12	lancopku/Avg-Avg [Findings of EMNLP 2022] Holistic Sentence Embeddings for Better...	30	Emerging	18	Python
13	INK-USC/RobustLR A Diagnostic Benchmark for Evaluating Logical Robustness of Deductive Reasoners	27	Experimental	8	Python
14	SecHack365-Fans/prompt2slip This library is testing the ethics of language models by using natural...	27	Experimental	9	Python
15	pramodkaushik/np_analysis Adversarial attacks generated for the ACL paper "Did the Model Understand...	26	Experimental	8	Jupyter Notebook
16	mirbostani/RobustQA RobustQA: A Framework for Adversarial Text Generation Analysis on Question...	26	Experimental	8	Python
17	snowood1/BERT-ENN Uncertainty-Aware Reliable Text Classification (KDD 2021)	24	Experimental	18	Jupyter Notebook
18	pramodkaushik/acl18_results Code to reproduce results in our ACL 2018 paper "Did the Model Understand...	24	Experimental	33	—
19	thunlp/RobTest Source code for ACL 2023 Findings paper "From Adversarial Arms Race to...	21	Experimental	8	Python
20	LCS2-IIITD/HIT-ACL2021-Codemixed-Representation This repo contains the source code of HIT: A Hierarchically Fused Deep...	20	Experimental	6	Python
21	Samanth-ai/robust-transformer-qa-adversarial-inoculation Transformer-based extractive QA robustness: adversarial distractor...	19	Experimental	—	Python
22	ruizheng20/robust_data Codes for the ACL Findings paper "Characterizing the Impacts of Instances on...	17	Experimental	5	Python
23	Kyle1668/LLM-TTA Code for the paper: Improving Black-box Robustness with In-Context Rewriting	16	Experimental	4	Jupyter Notebook
24	alekseevskaia/OrderBkd Backdoor attack in NLP	15	Experimental	2	Python
25	Efi-Pecani/Literary-LLM-Knowledge-Data-Poisoning Data poisoning attacks on LLMs — corrupting Harry Potter knowledge via...	15	Experimental	1	Jupyter Notebook
26	karanparekh14/ADTA5760--Group-4 Contaminated knowledge base of 150 academic PDFs (450 contaminants) for LLM...	14	Experimental	—	HTML
27	xinzhel/attack_alta Experiment for the paper published on ALTA2021: "Exploring the Vulnerability...	14	Experimental	1	Python
28	lancopku/DAN [Findings of EMNLP 2022] Expose Backdoors on the Way: A Feature-Based...	14	Experimental	12	Python
29	leonardodalinky/zero-shot-GLS [NAACL'24] ZGLS: Zero-shot Generative Linguistic Steganography	14	Experimental	9	Python
30	gallilmaimon/LUNATC This is the official implementation of "A Universal Adversarial Policy for...	14	Experimental	9	Python
31	YangLinyi/GLUE-X We leverage 14 datasets as OOD test data and conduct evaluations on 8 NLU...	14	Experimental	93	Python
32	claws-lab/multimodal-robustness Code and resources for EMNLP 2022 paper on 'Robustness of Fusion-based...	13	Experimental	10	Python
33	chong-z/nlp-second-order-attack [NAACL 2021] Code for "Double Perturbation: On the Robustness of Robustness...	13	Experimental	8	Python
34	REACT-NLP/tcab_generation Code to generate and extend the TCAB dataset.	12	Experimental	3	Python
35	princeton-nlp/rationale-robustness NAACL 2022: Can Rationalization Improve Robustness? https://arxiv.org/abs/2204.11790	12	Experimental	27	Python
36	nishiwen1214/AT_Papers Must-read papers on Adversarial training for neural networks！	12	Experimental	12	—
37	jrohsc/MSDT Code of the IEEE UV2022 accepted paper "MSDT: Masked Language Model Scoring...	12	Experimental	3	Python
38	lexicalcomputing/hamod a High Agreement Multi-lingual Outlier Detection dataset	12	Experimental	4	Python
39	95anantsingh/NYU-Attacking-Compressed-NLP Testing transferability of Adversarial samples across NLP models.	10	Experimental	1	Jupyter Notebook