Presents a byte-native LLM with bespoke tokenizer achieving 69-98% accuracy on malware family and architecture classification from raw bytes.
Megabyte: Predicting million-byte sequences with multiscale transformers, 2023
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2verdicts
UNVERDICTED 2representative citing papers
Kronecker Embeddings replace learned embedding tables with a deterministic byte-level character-position factorization and single projection, reducing parameters over 90% with reported gains in loss and robustness on language modeling tasks.
citing papers explorer
-
Kronecker Embeddings: Byte-Level Structured Token Representations for Parameter-Efficient Language Models
Kronecker Embeddings replace learned embedding tables with a deterministic byte-level character-position factorization and single projection, reducing parameters over 90% with reported gains in loss and robustness on language modeling tasks.