Crossword
Across
- 3. Table resizing in extension
- 7. Token collection size
- 8. BPE's main goal
- 12. Extending pre-trained models
- 14. Merge count in BPE
- 15. Models extended with tokens
- 16. Simple tokenization
Down
- 1. LLM input units
- 2. Balancing act factor
- 4. Advanced data integration
- 5. Advanced token optimization
- 6. Byte Pair Encoding
- 9. Alternative tokenizer
- 10. Popular tokenizer family
- 11. SentencePiece operates on
- 13. Token representation