AI Paper Atlas

Research map / Detection and segmentation / Open-set and few-shot detection

Beyond In-Domain Detection: SpikeScore for Cross-Domain Hallucination Detection

Yongxin Deng, Zhen Fang, Sharon Li, Ling Chen

ICLR 2026 · Detection and segmentation › Open-set and few-shot detection · approach: Domain-adaptive object detection · 2 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
use SpikeScore (max second-order difference over multi-turn trajectory) instead of single-turn SAPLMA scoreSAPLMA probe · Cross-domain average, Llama-3.1-8Bmean AUROC: 0.5764 -> 0.7860Improved
use SpikeScore instead of single-turn SEP scoreSEP probe · Cross-domain average, Llama-3.1-8Bmean AUROC: 0.5282 -> 0.7684Improved
SpikeScore vs ICR Probe, strongest cross-domain baselineICR Probe · Cross-domain average, Qwen3-14Bmean AUROC: 0.7435 -> 0.7874Improved
SpikeScore vs ICR Probe on smaller modelICR Probe · Cross-domain average, Llama-3.2-3Bmean AUROC: 0.7463 -> 0.7474No clear effect
use SAPLMA backbone instead of Perplexity for SpikeScore (ablation)SpikeScore · Backbone average, Llama-3.1-8Bmean AUROC: 0.7160 -> 0.7860Improved
use SAPLMA backbone instead of Reasoning score for SpikeScore (ablation)SpikeScore · Backbone average, Llama-3.1-8Bmean AUROC: 0.7595 -> 0.7860Improved
use SAPLMA backbone instead of In-Context Sharpness (ablation)SpikeScore · Backbone average, Llama-3.1-8Bmean AUROC: 0.7311 -> 0.7860Improved
use SAPLMA backbone instead of SEP for SpikeScore (ablation)SpikeScore · Backbone average, Qwen3-14Bmean AUROC: 0.7718 -> 0.7874Improved
train-based backbones (SEP) instead of training-free (Perplexity) under SpikeScore (ablation)SpikeScore · Backbone average, Qwen3-8Bmean AUROC: 0.6742 -> 0.7293Improved
use SpikeScore in retrieval-augmented pipeline vs ICR ProbeICR Probe · RAGTruth, Llama-3.1-8BAUROC: 0.7405 -> 0.8154Improved
use SpikeScore in RAG vs ICR ProbeICR Probe · TriviaQA RAG, Qwen3-8BAUROC: 0.7495 -> 0.8413Improved
use second-order difference instead of coefficient of variation to measure fluctuation (ablation)SpikeScore · CoQA-trained, Llama-3.1-8BAUROCImproved

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