SPARSEFIT: Few-shot Prompting with Sparse Fine-tuning for Jointly Generating Predictions and Natural Language Explanations.
Jesus SolanoOana-Maria CamburuPasquale MinerviniPublished in: CoRR (2023)
Keyphrases
- fine tuning
- natural language
- viable alternative
- fine tune
- video sequences
- natural language processing
- language processing
- knowledge representation
- sparse representation
- semantic analysis
- compressed sensing
- fine tuned
- shot boundary detection
- learning algorithm
- video indexing
- semantic interpretation
- natural language generation
- generating explanations
- sparse data
- question answering
- video data
- machine learning
- compressive sensing
- dialogue system
- video shots
- video content
- key frames
- image classification
- information extraction
- sparse codes
- shot change detection