Task knowledge can be compressed into an external factorized operator that a frozen LLM reads through a trained interface, and this representation outperforms text and matched soft-prefix, prefix, and LoRA baselines on most tested settings.
Chandra and Dexter C
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
KV-Skill: Forging Expertise in the Model's Native Language
Task knowledge can be compressed into an external factorized operator that a frozen LLM reads through a trained interface, and this representation outperforms text and matched soft-prefix, prefix, and LoRA baselines on most tested settings.