The D-Score is introduced, a simple spectral statistic computed from a single forward pass that is used as a hallucination score, classifying an input text as hallucinated when its D-Score is larger than a pre-defined quantity.
Abstract
Large Language Models can produce fluent text that is false, unsupported by the available evidence, or inconsistent with information that appears to be internally represented by the model. We study hallucination detection from the geometry of hidden activations and introduce the D-Score, a simple spectral statistic computed from a single forward pass. For a fixed model, layer, and tolerance parameter, the D-Score counts how many singular directions of the hidden activation matrix have singular values that remain close to the leading one. We use this quantity as a hallucination score, classifying an input text as hallucinated when its D-Score is larger than a pre-defined quantity. The motivation is that, when a model processes a text that conflicts with information available in its own internal state, the hidden representation may encode both the asserted content and some form of counter-evidence, uncertainty, correction, or lack of support; this can make the hidden trajectory spread across additional singular directions. We formalize this intuition through a lightweight spectral argument and evaluate the resulting detector on FAVA-Annotation and RAGTruth. The experiments indicate that the D-Score is a strong hidden-state signal for hallucination detection, while requiring no external verifier, no retrieval step, and no multiple generations.
This paper treats hallucination as a temporally extended span and detects it by sequence labeling: each token is scored from a 33-dimensional feature stream that fuses text statistics, Natural Language Inference entailment, and language model surprisal, with no access to model internals.
This work proposes AURORA, a novel hallucination detection framework that shifts the focus from static representations to the weight-gradient dynamics of LLMs, and achieves strong hallucination detection performance across four model families and four benchmark datasets.
Z. Zhang, Hainan Zhang, Zhiming Zheng· arXiv.org· 0 citations
Despite their widespread use, Large Language Models (LLMs) remain limited by a fundamental problem: the generation of plausible but false content, known as hallucinations. Most existing detection methods operate at the answer or sentence level, yet per-token detection is essential for localizing hallucinated spans and enabling fine-grained interventions. In this paper, we explore the use of the Mixture-of-Experts (MoE) paradigm to address this gap. In MoE architectures, a single forward pass activates a sparse subset of experts (i.e., distinct feedforward networks per layer) via a routing mechanism, producing internal signals (e.g., router entropy, expert disagreement, and expert usage patterns) that are unavailable in dense architectures and have not been previously exploited for hallucination detection. To this end, we introduce InnerExpert, the first method to leverage these MoE-specific signals for per-token hallucination detection. InnerExpert combines routing-level and standard transformer signals into compact per-token feature vectors, classified by a lightweight detector trained on labels produced by an LLM-as-a-judge pipeline, which enables continuous model updates without manual annotation. Our results show that InnerExpert outperforms existing methods across five datasets and two MoE architectures, achieving up to 0.91 answer-level and 0.76 token-level AUROC, while requiring only a single forward pass.
João Fonseca, Rodrigo Rodrigues, Paolo Romano· 0 citations
TRE is a parameter-free and single-run metric that estimates hallucination risk directly from the entropy signals of a single generation, without requiring any detector training or repeated sampling, and enjoys strong generalizability, efficiency, and robustness.
Pengcheng Weng, Y. Qian, Yue Tan et al.· 0 citations
A lightweight linear detector is built on top of Role-Break that requires no fine-tuning of the VLM, whose feature dimension stays below 5,000 and reaches an average AUROC of 93.23 across six VLMs and four benchmarks.
Mingyu Wang, Weilin Jin, Wenbo Li et al.· 0 citations
ReWEIGH is a training-free decoding intervention that aggregates vocabulary ranks across visual positions and compares each candidate with a token-specific reference estimated from unlabeled images and applies a bounded penalty only to candidates that fall below their reference.