2026· Annual Meeting of the Association for Computational Linguistics· pp. 21786-21800· 0 citations· 33 references
Computer Science
TL;DR
GHOST is proposed, an efficient white-box framework for hallucination detection in LLMs that primarily target confused hallucinations marked by internal reasoning instability, while also capturing stub-born hallucinations characterized by premature layer-wise convergence as a complementary signal.
Abstract
Large Language Models (LLMs) often generate overconfident yet factually incorrect hallucinations. Current detection paradigms suffer from a trade-off between the high accuracy of computationally expensive black-box methods and the inability of white-box methods to detect stubborn hallucinations. To bridge this gap, we propose GHOST ( G eometric H idden-state O bservation for S emantic T ruthfulness), an efficient white-box framework for hallucination detection in LLMs. We primarily target confused hallucinations marked by internal reasoning instability, while also capturing stub-born hallucinations characterized by premature layer-wise convergence as a complementary signal. By integrating internal geometric dynamics with output probability distributions, GHOST constructs a high-dimensional feature space for non-linear truthfulness classification. Extensive evaluations on FinanceBench, RAGTruth, HaluEval, and PopQA show that GHOST outperforms white-box baselines and achieves competitive black-box performance while reducing computational overhead by over 90%, offering a robust solution for real-time detection.
HalluTracer is introduced, a detection framework that reads and aggregates truthfulness evidence across every layer of the forward pass before the model emits any answer token, recasts hallucination detection from a layer-selection problem into a depth-aggregation problem governed by the geometric sparsity of the truthfulness signal.
Zhihao Guo, Zonghan Wu, Huan Huo et al.· 0 citations
HallDetect, a lightweight, reference-free, and black-box framework for hallucination detection, is presented, a lightweight, reference-free, and black-box framework for hallucination detection that is evaluated not only on summarization but across a broader range of source-grounded generation settings.
Achir Oukelmoun, N. Semmar, Gäel de Chalendar· 0 citations
TruthLens is a self-evaluation framework that teaches the LM head to expose a per-object truthfulness signal without any auxiliary model or additional inference cost, and generalizes effectively to benchmarks with substantially larger label spaces.
Yanqi Wu, Runhe Lai, Xinhua Lu et al.· 0 citations
Results on the object hallucination evaluation benchmark POPE show that the proposed Dual-branch Inference for Visual-prior Elimination reduces the model’s hallucination rate when objects are dense or complex occlusions exist, and brings a slight improvement in the question-answering accuracy of the model on the MSCOCO and VG datasets.
Shuguo Jiang· Journal of Electronics and I...· 0 citations
Visual counting is a fundamental pillar of multimodal intelligence, requiring a seamless integration of fine-grained grounding and spatial reasoning. While Multimodal Large Language Models (MLLMs) have achieved remarkable success in qualitative scene understanding, their quantitative precision remains a significant bottleneck, often characterized by persistent numerical hallucinations. Existing counting benchmarks primarily focus on basic perception in simplified contexts, failing to capture the complex failure modes that emerge under logical constraints or adversarial conditions. To address these limitations, we introduce HoloCount, a holistic and diagnostically rich benchmark structured around a three-level hierarchical taxonomy. HoloCount evaluates MLLMs across: (1) Semantic Counting, focusing on atomic and property-based enumeration; (2) Analytical Counting, assessing logical composition through spatial and set-based reasoning; and (3) Robustness Testing, probing model integrity against adverse scenarios and grounded counter-priors, such as high-density scenes and linguistic biases. Through an exhaustive evaluation of over 20 state-of-the-art MLLMs, we reveal a critical performance gap: even top-tier models degrade significantly as tasks transition from perception to complex analytical reasoning and adverse scenarios. Our findings provide a systematic landscape of current MLLM counting capabilities and offer a roadmap for developing more grounded and reliable multimodal systems. The dataset is available at https://mm-mvr.github.io/HoloCount/.
A novel inference-time intervention method that encodes truthful and untruthful content as an entailment hierarchy and performs representation editing in the hyperbolic space to activate the truthfulness of LLMs is proposed.
Tongxu Lin, Junping Du, Zhe Xue et al.· Proceedings of the 32nd ACM...· 0 citations