Unverified50% confidenceFactExact time
GPTQ was proposed by Frantar et al. in 2022 and minimizes quantization error based on the inverse of the Hessian matrix; AWQ was proposed by Ji Lin et al. at MIT, with the core insight that a small number of weight channels associated with large activation values should retain higher precision.
1
Sources
50%
Confidence
Long-term
Relevance
9/2/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedGPTQ（2022）引入逐层Hessian矩阵近似，通过最小化权重量化前后的输出重建误差优化量化参数75% similarUnverifiedThe GPTQ quantization algorithm was proposed by Tim Dettmers et al. in 2023, optimizing quantization error layer by layer through second-order gradient information.74% similarUnverified现代量化方法如 GPTQ、AWQ 通过逐层校准显著缓解了精度退化问题72% similarUnverifiedGPTQ was proposed in 2022 and extends the OBS framework from pruning to quantization, using the inverse of the Hessian matrix to optimally adjust unquantized weights to compensate for quantization error, with efficient updates via Cholesky decomposition.71% similarUnverifiedAWQ由MIT韩松团队于2023年提出，通过保护显著权重，相较朴素Q4量化在相同压缩比下可减少约20%的精度损失70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/845847API
curl https://kongchang.com/api/v1/knowledge/claims/845847MCP
get_claim(id=845847)