AI Paper Atlas

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

From Out-of-Distribution Detection to Hallucination Detection: A Geometric View

Litian Liu, Reza Pourreza, Yubing Jian, Yao Qin, Roland Memisevic

ICML 2026 · Large language models › Language, safety and interpretability · approach: Data and hallucination detection · 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
use analytical decision-neutral point as feature mean instead of empirical training mean (ablation)NCI · CSQA (Llama-3.2-3B-Instruct)AUROC: 62.79 -> 66.07Improved
use NCI with analytical proxy instead of Perplexityhallucination detector · CSQA (Llama-3.2-3B-Instruct)AUROC: 63.23 -> 66.07Improved
restrict fDBD to top-1000 alternative tokens instead of all tokens (ablation)fDBD · CSQA (Llama-3.2-3B-Instruct)AUROC: 68.15 -> 69.24Improved
restrict fDBD to top-100000 alternative tokens instead of 1000 (ablation)fDBD · CSQA (Llama-3.2-3B-Instruct)AUROC: 69.24 -> 68.26Hurt
restrict fDBD to top-1 alternative token (ablation)fDBD · CSQA (Llama-3.2-3B-Instruct)AUROC: 69.24 -> 68.64No clear effect
perturb selected k by 20% (ablation)fDBD · GSM8K (Llama-3.2-3B-Instruct)AUROC: 76.36 -> 76.32No clear effect
use fDBD with selected k instead of all tokens (ablation)fDBD · GSM8K (Llama-3.2-3B-Instruct)AUROC: 75.59 -> 76.36Improved
use fDBD with selected k instead of all tokens (ablation)fDBD · CSQA (Qwen-2.5-7B-Instruct)AUROC: 71.50 -> 72.47Improved
apply stochastic decoding at temperature 1.0 instead of 0.2fDBD · CSQA (Llama-3.2-3B-Instruct)AUROC: 69.30 -> 69.19No clear effect
apply stochastic decoding at temperature 1.0 instead of 0.2NCI · CSQA (Llama-3.2-3B-Instruct)AUROC: 67.07 -> 67.93No clear effect
use fDBD instead of Perplexity on larger modelfDBD (Qwen-3-32B) · GSM8KAUROC: 76.86 -> 80.60Improved
use fDBD instead of NCI on MoE modelfDBD (Qwen-3-30B-A3B) · AQuAAUROC: 70.09 -> 69.60Hurt

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