A multi-pass prompt verification method stabilizes and improves accuracy of 4-bit and lower quantized LLaMA-3.1 models for thematic extraction from interview data, though 8-bit versions remain closest to a human-plus-BF16 gold standard.
In: Findings of NAACL 2025, pp
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Improving Quantized Model Performance in Qualitative Analysis with Multi-Pass Prompt Verification
A multi-pass prompt verification method stabilizes and improves accuracy of 4-bit and lower quantized LLaMA-3.1 models for thematic extraction from interview data, though 8-bit versions remain closest to a human-plus-BF16 gold standard.