A graph neural network trained on Hessian-derived graphs of LLM weights adaptively assigns per-column quantization bit-widths, improving low-bit perplexity over GPTQ on LLaMA models.
Measuring massive multitask language understanding,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
Mixed-Precision Graph Neural Quantization for Low Bit Large Language Models
A graph neural network trained on Hessian-derived graphs of LLM weights adaptively assigns per-column quantization bit-widths, improving low-bit perplexity over GPTQ on LLaMA models.