Research Papers 论文研究 8d ago Updated 7d ago 更新于 7天前 43

Reliability-Aware Sexism Detection: Combining DPO with Annotator Agreement and Token-Level Confidence Scoring 可靠性感知性别歧视检测:结合DPO、标注者一致性和Token级置信度评分

RA-DPO integrates annotator agreement, model confidence, and token-level uncertainty into a single reliability score for preference optimization Training on only the top 30% most reliable preference pairs matches full-data DPO performance, demonstrating significant cost reduction Selective prediction achieves 96.2% accuracy at 50% coverage (true-agreement) and 88.7% (predicted-agreement), both surpassing the 85.3% no-agreement baseline The approach enables inference-time abstention, trading cove 提出RA-DPO方法,整合标注者一致性、模型置信度和token级不确定性信号构建可靠性评分 在EXIST 2023数据集(6,920条多语言帖子)上验证,训练前30%最可靠偏好对即可匹配完整数据DPO性能 推理时选择性预测在50%覆盖率下达到96.2%准确率(真实一致性)和88.7%(预测一致性),均优于85.3%基线 使用OpenAI gpt-4o base进行DPO微调,并在Llama和Qwen两个3B开源模型上验证

55
Hot 热度
72
Quality 质量
60
Impact 影响力

Analysis 深度分析

TL;DR

  • RA-DPO integrates annotator agreement, model confidence, and token-level uncertainty into a single reliability score for preference optimization
  • Training on only the top 30% most reliable preference pairs matches full-data DPO performance, demonstrating significant cost reduction
  • Selective prediction achieves 96.2% accuracy at 50% coverage (true-agreement) and 88.7% (predicted-agreement), both surpassing the 85.3% no-agreement baseline
  • The approach enables inference-time abstention, trading coverage for accuracy in subjective classification tasks
  • Evaluated on 6,920 multilingual posts from EXIST 2023 with fine-tuning of gpt-4o and validation on Llama and Qwen 3B models

Why It Matters

This work addresses a fundamental limitation in subjective NLP classification: the loss of annotation uncertainty when collapsing multi-annotator labels into majority decisions. For AI practitioners building safety or content moderation systems, RA-DPO offers a practical framework to improve reliability without additional labeling costs, while providing a deployable abstention mechanism that can reduce harmful false positives in high-stakes applications.

Technical Details

  • RA-DPO (Reliability-Aware Direct Preference Optimization) combines three signals—annotator agreement, model confidence, and token-level uncertainty—into a unified reliability score used to rank preference pairs during DPO training
  • Data: 6,920 multilingual posts from the EXIST 2023 shared task on sexism detection, with multiple annotator labels per instance
  • Training: Fine-tuned OpenAI gpt-4o base model via DPO; validated generalization on two open-weight 3B models (Llama and Qwen)
  • Selective prediction: At inference, the model can abstain on low-reliability samples, achieving 96.2% accuracy at 50% coverage under true-agreement conditions and 88.7% under the more realistic predicted-agreement setting, compared to 85.3% for a baseline ignoring agreement
  • Key finding: Training exclusively on the top 30% most reliable preference pairs achieves performance parity with full-data DPO, indicating that reliability-aware filtering can substantially reduce training compute and cost

Industry Insight

  • Content moderation and safety classification systems should move beyond majority-vote labels and incorporate annotator disagreement as a first-class signal, particularly for subjective categories like sexism, hate speech, and harassment
  • Selective prediction with abstention is a deployable strategy for reducing false positives in production moderation pipelines, where the cost of erroneous decisions often outweighs the cost of missed detections
  • Reliability-aware data selection during preference optimization could become a standard technique for efficient fine-tuning across subjective NLP tasks, offering both cost savings and improved deployment safety

TL;DR

  • 提出RA-DPO方法,整合标注者一致性、模型置信度和token级不确定性信号构建可靠性评分
  • 在EXIST 2023数据集(6,920条多语言帖子)上验证,训练前30%最可靠偏好对即可匹配完整数据DPO性能
  • 推理时选择性预测在50%覆盖率下达到96.2%准确率(真实一致性)和88.7%(预测一致性),均优于85.3%基线
  • 使用OpenAI gpt-4o base进行DPO微调,并在Llama和Qwen两个3B开源模型上验证

为什么值得看

本文针对在线性别歧视检测这一主观性强的NLP任务,提出了可靠性感知的训练与推理框架,为标注不确定性建模提供了新思路。对AI从业者而言,该方法展示了如何在主观分类任务中平衡覆盖率与准确率,具有实际应用价值。

技术解析

  • RA-DPO方法:将标注者一致性、模型置信度和token级不确定性信号整合为单一可靠性评分,用于训练时选择高价值偏好对和推理时的拒绝预测机制
  • 实验设置:基于EXIST 2023数据集的6,920条多语言帖子,使用OpenAI gpt-4o base进行DPO微调,并在Llama和Qwen两个3B开源模型上验证
  • 训练效率:仅使用Top 30%最可靠偏好对进行训练,性能与完整数据DPO相当,显著降低训练成本
  • 推理性能:选择性预测在50%覆盖率下达到96.2%准确率(真实一致性设置)和88.7%(可部署的预测一致性设置),均超过85.3%无一致性基线

行业启示

  • 主观性NLP任务(如仇恨言论检测、内容审核)应重视标注不确定性建模,而非简单采用多数投票
  • 可靠性感知的数据筛选策略可在保持性能的同时显著降低训练成本,适合资源受限场景
  • 推理时的选择性预测机制为高风险应用场景提供了"覆盖-准确率"权衡的实用方案

Disclaimer: The above content is generated by AI and is for reference only. 免责声明:以上内容由 AI 生成,仅供参考。

Fine-tuning 微调 Alignment 对齐 Evaluation 评测 Research 科学研究 Dataset 数据集