AI Paper Atlas

Research map / Large language models / Agents, code and math

HalluGuard: Demystifying Data-Driven and Reasoning-Driven Hallucinations in LLMs

Xinyue Zeng, Junhong Lin, Yujun Yan, Feng Guo, Liang Shi, Jun Wu, Dawei Zhou

ICLR 2026 · Large language models › Agents, code and math · approach: Mathematical and structural reasoning benchmarks · 8 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 HalluGuard score instead of Insidedetector · RAGTruth (QwQ-32B)AUROC: 77.72 -> 84.59Improved
use HalluGuard score instead of Insidedetector · MATH-500 (QwQ-32B)AUROC: 80.8 -> 81.76Improved
use HalluGuard score instead of RACE/Insidedetector · TruthfulQA (Mistral-7B)AUROC: 71.06 -> 77.05Improved
use HalluGuard on smaller backbonedetector · HaluEval (Llama2-7B)AUPRC: 72.89Improved
use HalluGuard on Llama2-13Bdetector · GSM8KAUROC: 75.79 -> 79.01Improved
use HalluGuard on Llama2-70Bdetector · SQuADAUROC: 81.24 -> 83.8Improved
use HalluGuard on Llama2-70Bdetector · HaluEvalAUROC: 74.62 -> 80.79Improved
guide beam search with HalluGuard score instead of IO PromptQwen2.5-Math-7B · MATH-500accuracy: 72.70 -> 81.00Improved
guide beam search with HalluGuard score instead of IO PromptLlama3.1-8B · Naturalaccuracy: 55.24 -> 70.96Improved
use HalluGuard for fine-grained semantic hallucinationdetector · PAWS (QwQ-32B)AUROC: 91.24Improved
use HalluGuard on GPT-2detector · PAWSAUROC: 83.27Improved
add the -log kappa^2 stability term (ablation)HalluGuard score · MATH-500 (increasing reasoning drift)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