A new method sheds light on how NLP models make decisions.
― 6 min read
Hassan Sajjad
A new benchmark evaluates language models' understanding of word meanings and relationships.
― 5 min read
This study evaluates how model size and quantization impact language model performance.
― 7 min read
A study introduces RepNoise to protect large language models from harmful fine-tuning.
― 6 min read
A dataset to test language models' grasp of wording differences.
― 5 min read
A novel approach to improve multimodal learning with missing data.
― 5 min read
A new single-branch method improves machine learning performance with missing data.
― 5 min read
A new method to assess language model outputs using multiple LLM judges.
― 7 min read
PENME enhances language model updates, tackling lexical bias and computational efficiency.
― 6 min read
Exploring Reverse Preference Attacks and their impact on model safety.
― 5 min read
Vision-Language Models face challenges in understanding language structure for image-text tasks.
― 6 min read