Papers with Code 아카이브
Adversarial Text
태스크의 하위 분류와 논문·코드·평가·기법·데이터셋을 탐색하세요.
- 하위 태스크
- 0
- 평가표
- 0
- 보존된 평가 기록
- 0
Adversarial Text refers to a specialised text sequence that is designed specifically to influence the prediction of a language model. Generally, Adversarial Text attack are carried out on Large Language Models (LLMs). Research on understanding different adversarial approaches can help us build effective defense mechanisms to detect malicious text input and build robust language models.
논문
이 태스크와 하위 분야의 관련 논문입니다. 실제 성능 비교는 평가 결과에서 확인하세요.
RETSim: Resilient and Efficient Text Similarity
BERT Lost Patience Won't Be Robust to Adversarial Slowdown
Towards a Robust Detection of Language Model Generated Text: Is ChatGPT that Easy to Detect?
VoteTRANS: Detecting Adversarial Text without Training by Voting on Hard Labels of Transformations
How do humans perceive adversarial text? A reality check on the validity and naturalness of word-based adversarial attacks
Iterative Adversarial Attack on Image-guided Story Ending Generation
Less is More: Removing Text-regions Improves CLIP Training Efficiency and Robustness
Towards Imperceptible Document Manipulations against Neural Ranking Models
A Pilot Study of Query-Free Adversarial Attack against Stable Diffusion
Frauds Bargain Attack: Generating Adversarial Text Samples via Word Manipulation Process
Improved Training of Mixture-of-Experts Language GANs
RETVec: Resilient and Efficient Text Vectorizer
TextDefense: Adversarial Text Detection based on Word Importance Entropy
Step by Step Loss Goes Very Far: Multi-Step Quantization for Adversarial Text Attacks
A survey on text generation using generative adversarial networks
Ignore Previous Prompt: Attack Techniques For Language Models
TAPE: Assessing Few-shot Russian Language Understanding
Adversarial Text Normalization
SemAttack: Natural Textual Attacks via Different Semantic Spaces
"That Is a Suspicious Reaction!": Interpreting Logits Variation to Detect NLP Adversarial Attacks
Adversarial Robustness of Neural-Statistical Features in Detection of Generative Transformers
Data-Driven Mitigation of Adversarial Text Perturbation
Identifying Adversarial Attacks on Text Classifiers
Repairing Adversarial Texts through Perturbation
Generating Watermarked Adversarial Texts
SEPP: Similarity Estimation of Predicted Probabilities for Defending and Detecting Adversarial Text
Don't Search for a Search Method -- Simple Heuristics Suffice for Adversarial Text Attacks
Semantic-Preserving Adversarial Text Attacks
DISCO : efficient unsupervised decoding for discrete natural language problems via convex relaxation
MATE-KD: Masked Adversarial TExt, a Companion to Knowledge Distillation
"TL;DR:" Out-of-Context Adversarial Text Summarization and Hashtag Recommendation
Adversarial Text-to-Image Synthesis: A Review
Persistent Anti-Muslim Bias in Large Language Models
Generating Natural Language Attacks in a Hard Label Black Box Setting
From Unsupervised Machine Translation To Adversarial Text Generation
출처: Papers with Code · CC BY-SA 4.0. 아카이브 자료와 새로 수집한 논문은 별도로 관리합니다.