AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

Detecting Contextual Hallucinations in Large Language Models with Frequency-Aware Attention

Siya Qi, Yudong Chen, Runcong Zhao, Qinglin Zhu, Zhanghao Hu, Wei Liu, Yulan He, Zheng Yuan, Lin Gui

ICML 2026 · Large language models › Language, safety and interpretability · approach: Task vectors and steering · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
extract high-frequency attention components instead of using full-spectrum or low-pass features (ablation)Fourier attention features · RAGTruth and HalluRAG (avg AUROC)AUROCImproved
use low-pass or full-spectrum attention features instead of high-pass (ablation)Fourier attention features · RAGTruth and HalluRAG (avg AUROC)AUROCNo clear effect
replace attention variance/entropy summaries with frequency-aware high-frequency energyattention-based detector · RAGTruth and HalluRAGAvg AUROC: 0.7627 -> 0.8052 (LLaMA-7B, entropy vs Fourier-high)Improved
use Fourier-high instead of Lookback-Lens on summarizationhallucination detector · RAGTruth-Summ (LLaMA-13B)AUROC: 0.6702 -> 0.7362Improved
use Fourier-high instead of Lookback-Lens on Mistral-7B QAhallucination detector · RAGTruth-QA (Mistral-7B)AUROC: 0.9148 -> 0.9099No clear effect
use Fourier operator instead of Wavelet or Laplacian (ablation)frequency-aware features · RAGTruth + HalluRAGAvg AUROC: 0.8052 / 0.7953 / 0.7859 (LLaMA-7B)Mixed
restrict the classifier to the top 100 heads (ablation)Fourier-high · RAGTruth-Avg (LLaMA-7B)AUROC: 0.8205 -> 0.7915Hurt
restrict the classifier to the top 10 heads (ablation)Fourier-high · RAGTruth-Avg (LLaMA-7B)AUROC: 0.8205 -> 0.6995Hurt
remove context-attention features (ablation)frequency-aware detector · RAGTruth and HalluRAGAUROCHurt
aggregate features into 8-token spans instead of single tokensFourier-high · RAGTruth-Avg (LLaMA-7B)AUROC: 0.8052 -> 0.8412 vs token-level avg (different task sets)Improved
use Fourier-high instead of Lookback-Lens in cross-task transferhallucination detector · cross-domain (train on one task, test on another)AUROCImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy