SkillWeave partitions LLM capabilities into compressible skillpacks to deliver strong multi-domain performance with a 9B model that outperforms larger monolithic LLMs and achieves up to 4x speedup on benchmarks.
We rescale: X←X· diag(s−1), W←diag(s)·W thus transferring channel-wise quantization difficulty from X to W
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Skill Weaving: Efficient LLM Improvement via Modular Skillpacks
SkillWeave partitions LLM capabilities into compressible skillpacks to deliver strong multi-domain performance with a 9B model that outperforms larger monolithic LLMs and achieves up to 4x speedup on benchmarks.