Files
CompanionGuard-RL/paper/sections/07_experiments.tex
zhangsiyuan 804ebd2f77 feat: add paper/ LaTeX draft, English data scripts, update progress docs
- paper/: 22-page LaTeX framework (7/10 sections complete, compiles cleanly)
  main.tex + 10 section files + refs.bib + compiled PDF (329KB)
- code/scripts/: three English dataset generation & merging scripts
  generate_english.py / generate_english_targeted.py / merge_v5.py
- CLAUDE.md: update paper writing status, add paper/ file map entry
- state.md: add section 8 paper writing progress (2026-05-15)
- .gitignore: add LaTeX build artifact exclusion rules

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
2026-05-18 11:19:39 +08:00

69 lines
2.6 KiB
TeX
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

% ============================================================
\section{实验}
\label{sec:experiments}
% ============================================================
\subsection{实验设置}
\subsubsection{评测集}
所有实验均在CompanionRisk-Bench测试集$n=1,486$)上进行。
为验证泛化性Module B的评估额外在non-homogeneous子集
393条真实人-AI对话上进行独立报告。
\subsubsection{评测指标}
\textbf{检测任务Module B}
\begin{itemize}
\item Binary F1有风险/无风险二分类F1
\item High-risk Recall高风险样本$y_\text{risk}=1$的召回率)
\item False Negative Rate (FNR)(漏检率)
\item Level Weighted F1风险等级5分类加权F1
\item Fine Macro F114类细粒度标签宏平均F1
\end{itemize}
\textbf{干预任务Module C}
\begin{itemize}
\item Safety RecallL3/L4高风险样本被正确干预比例
\item Over-refusal RateL0安全样本被REWRITE及以上干预的比例
\item Action Accuracy与标注推荐动作$a_\text{recommend}$的吻合率)
\item Crisis PrecisionCRISIS动作中L4样本的比例
\item Safety-UX F-score安全召回率与过拒率的调和平均衍生得分
\end{itemize}
\subsubsection{基线方法}
\textbf{检测基线}
L1a关键词匹配、L1b正则词典、L1c组合
\todo{L2Llama Guard v2、WildGuard、OpenAI Moderation待运行}
\textbf{干预基线}
Rule-based$l_\text{risk} \geq 3$即REJECT其余PASS
Threshold Baseline按风险分数阈值映射动作
\todo{LLM-as-judgeQwen2.5-72B直接判断待运行}
\subsection{RQ1检测性能分析}
详细结果见第\ref{sec:moduleB}节表\ref{tab:moduleB_main}和表\ref{tab:per_category_recall}
Module B在所有指标上大幅优于基线。
值得关注的是,通用守卫模型(\todo{Llama Guard v2、WildGuard}
在伴侣特有风险类别R3情感操纵、R4现实隔离等上的召回率
预期显著低于整体水平,
体现了CompanionRisk Taxonomy的必要性。
\subsection{RQ2干预策略比较}
\todo{本节主要结果待Module C v5完成后填入。}
核心发现基于v3结果
RL策略在safety\_recall1.0 vs 0.908)和
UX F-score0.998 vs 0.952)上均优于两个基线策略,
证明了可学习干预策略相比固定规则的优越性。
\subsection{RQ3消融实验}
\todo{消融实验表格待补充。预期包含:
(1) Module BResponse-only / History+R / Persona+R / Full
(2) Module CBC-only / RL w/o category reward / Full RL。}