
-
Calibrate Before Use: Improving Few-Shot Performance of Language Models
GPT-3 can perform numerous tasks when provided a natural language prompt...
read it
-
Extracting Training Data from Large Language Models
It has become common to publish large (billion parameter) language model...
read it
-
AutoPrompt: Eliciting Knowledge from Language Models with Automatically Generated Prompts
The remarkable success of pretrained language models has motivated the s...
read it
-
Customizing Triggers with Concealed Data Poisoning
Adversarial attacks alter NLP model predictions by perturbing test-time ...
read it
-
Gradient-based Analysis of NLP Models is Manipulable
Gradient-based analysis methods, such as saliency map visualizations and...
read it
-
Trustworthy AI Inference Systems: An Industry Research View
In this work, we provide an industry research view for approaching the d...
read it
-
Imitation Attacks and Defenses for Black-box Machine Translation Systems
We consider an adversary looking to steal or attack a black-box machine ...
read it
-
Pretrained Transformers Improve Out-of-Distribution Robustness
Although pretrained Transformers such as BERT achieve high accuracy on i...
read it
-
Evaluating NLP Models via Contrast Sets
Standard test sets for supervised learning evaluate in-distribution gene...
read it
-
Train Large, Then Compress: Rethinking Model Size for Efficient Training and Inference of Transformers
Since hardware resources are limited, the objective of training deep lea...
read it
-
AllenNLP Interpret: A Framework for Explaining Predictions of NLP Models
Neural NLP models are increasingly accurate but are imperfect and opaque...
read it
-
Do NLP Models Know Numbers? Probing Numeracy in Embeddings
The ability to understand and work with numbers (numeracy) is critical f...
read it
-
Universal Adversarial Triggers for Attacking and Analyzing NLP
Adversarial examples highlight model vulnerabilities and are useful for ...
read it
-
Universal Adversarial Triggers for NLP
Adversarial examples highlight model vulnerabilities and are useful for ...
read it
-
Compositional Questions Do Not Necessitate Multi-hop Reasoning
Multi-hop reading comprehension (RC) questions are challenging because t...
read it
-
Misleading Failures of Partial-input Baselines
Recent work establishes dataset difficulty and removes annotation artifa...
read it
-
Understanding Impacts of High-Order Loss Approximations and Features in Deep Learning Interpretation
Current methods to interpret deep learning models by generating saliency...
read it
-
Interpreting Neural Networks With Nearest Neighbors
Local model interpretation methods explain individual predictions by ass...
read it
-
Trick Me If You Can: Adversarial Writing of Trivia Challenge Questions
Modern natural language processing systems have been touted as approachi...
read it
-
Right Answer for the Wrong Reason: Discovery and Mitigation
Exposing the weaknesses of neural models is crucial for improving their ...
read it