八位重建
More quantization levels usually reduce reconstruction error.
本地计算 / 模拟本地教学实验，不调用外部 API。合成示例不能替代生产环境测量。
学习指南与本课目录
LEARNING CONTRACT理解 → 实现 → 研究
目标：解释本课机制、复现一个例子、测试一个失败案例。读完不自动代表掌握。
- 理解层：写明假设、形状/单位和一个反例。
- 实现层：复算数值例子并检查边界。
- 研究层：在匹配预算下比较一个替代方法。
02 · 预测与配置
技术场景说明、参数名称与参考答案目前为英文。
03 · 检查证据
方法与参考资料
Symmetric per-tensor scalar quantizer with range ±(2^(bits−1)−1). This is not LLM.int8(), GPTQ or AWQ. Ideal compression ignores scale metadata and runtime packing. Scalar error does not measure language-model accuracy.
Text / 原始公式
scale=max|x|/(2^(bits−1)−1); q=round(x/scale); x̂=q×scale
Low scalar MSE does not guarantee stable downstream model quality.
LLM.int8() · outliers and mixed precision ↗练习题
quantization-2 / 01What does “Eight-bit reconstruction” test?
More quantization levels usually reduce reconstruction error.
Original teaching answer / local derivation. Method reference: LLM.int8() · outliers and mixed precision ↗
打开实验 →quantization-2 / 02Which computation produces the result in “Eight-bit reconstruction”?
Original teaching answer / local derivation. Method reference: LLM.int8() · outliers and mixed precision ↗
打开实验 →quantization-2 / 03What is the baseline configuration for “Eight-bit reconstruction”?
{"values":"-1,-0.7,-0.2,0,0.15,0.4,0.8,1","bits":8}
Original teaching answer / local derivation. Method reference: LLM.int8() · outliers and mixed precision ↗
打开实验 →quantization-2 / 04What numerical result should you check for “Eight-bit reconstruction”?
{"MSE":0.0000028868807737615443,"Max absolute error":0.003149606299212626,"Ideal compression vs FP32":4,"Scale":0.007874015748031496}
Original teaching answer / local derivation. Method reference: LLM.int8() · outliers and mixed precision ↗
打开实验 →quantization-2 / 05How would you run a controlled comparison for “Eight-bit reconstruction”?
Keep the supplied fixture and other settings fixed; change bits. Compare reconstruction error, MSE, scale and ideal compression.
Original teaching answer / local derivation. Method reference: LLM.int8() · outliers and mixed precision ↗
打开实验 →quantization-2 / 06What failure could invalidate a conclusion from “Eight-bit reconstruction”?
Low scalar MSE does not guarantee stable downstream model quality.
Original teaching answer / local derivation. Method reference: LLM.int8() · outliers and mixed precision ↗
打开实验 →quantization-2 / 07Is “Eight-bit reconstruction” a benchmark of a deployed language model?
No. Symmetric per-tensor scalar quantizer with range ±(2^(bits−1)−1). This is not LLM.int8(), GPTQ or AWQ. Ideal compression ignores scale metadata and runtime packing. Scalar error does not measure language-model accuracy.
Original teaching answer / local derivation. Method reference: LLM.int8() · outliers and mixed precision ↗
打开实验 →quantization-2 / 08Which evidence should you report for “Eight-bit reconstruction”?
Compare reconstruction error, MSE, scale and ideal compression. Export inputs, engine, result and method; distinguish synthetic fixtures from observed production data.
Original teaching answer / local derivation. Method reference: LLM.int8() · outliers and mixed precision ↗
打开实验 →quantization-2 / 09How can another researcher reproduce “Eight-bit reconstruction”?
Open quantization-2, reset the preset, retain the exact inputs and run. The calculation is deterministic for identical inputs.
Original teaching answer / local derivation. Method reference: LLM.int8() · outliers and mixed precision ↗
打开实验 →quantization-2 / 10What is the next validation step after “Eight-bit reconstruction”?
Low scalar MSE does not guarantee stable downstream model quality. Test the hypothesis on a representative held-out workload and document a baseline, success metric and stopping rule before claiming deployment readiness.
Original teaching answer / local derivation. Method reference: LLM.int8() · outliers and mixed precision ↗
打开实验 →假设、参考实现与验收
这是对称逐张量舍入；理想压缩率不含元数据。全零张量安全尺度取 1。标量重建 MSE 不是 LLM 质量。
已复核引擎公式、场景预设及数值保护。固定种子的合成输入不是生产基准。
运行前：预测一个边界情况
More quantization levels usually reduce reconstruction error.
提示：隔离假设
固定基线输入，只改变一个参数或边界；同时比较数值与声明单位。单例通过仅验证该样例。
最小可运行算例
将算例及所链接模块保存在同一目录。使用 Node.js 22 或更高版本，然后运行 node quantization-2.mjs. 无需第三方运行依赖。
import {calculate,network} from './academy-core.mjs';
const params={
"values": "-1,-0.7,-0.2,0,0.15,0.4,0.8,1",
"bits": 8
};
const result=calculate('quantization',params);
console.log(result.metrics);测试与迁移
More quantization levels usually reduce reconstruction error.
在下方笔记记录输入、代码版本、期望值和实测值；再构造违反上述假设的反例，解释原结论为何不再成立。
下载完整代码包（.tar.gz） · 完整参考代码与回归检查预测、检验与证据记录
展开答案前先预测一个结果；练习后记录差异和迁移测试。记录仅存本地，在不同语言版本间共用。
打开记录／导出／清除 →数学复核范围逐篇记录，不代表完成论文实证复现；可在笔记中记录疑似错误。 复核范围