Detecting Hallucinations in Retrieval-Augmented Generation through Grounding-Aware Sensitivity by Perturbation (GASP)

• URL: http://arxiv.org/abs/2607.04223v1
• Date: Sun, 05 Jul 2026 10:35:30 GMT
• Title: Detecting Hallucinations in Retrieval-Augmented Generation through Grounding-Aware Sensitivity by Perturbation (GASP)
• Abstract summary: Grounding-Aware Sensitivity by Perturbation (GASP) is a span-level detector that scores each answer sentence by how strongly its likelihood depends on the retrieved evidence.<n>GASP holds the answer fixed and re-scores it under the full context, under no context, and with each chunk removed, then measures the log-likelihood drops and Jensen-Shannon divergences.<n>We evaluate GASP on three benchmarks (RAGTruth, TofuEval, RAGBench) with three instruction-tuned scorers from two model families.
• Score: 1.827510863075184
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Retrieval-augmented generation (RAG) reduces but does not eliminate hallucination, and existing detectors return a single answer-level score that does not indicate which sentence is unsupported, or why. To close this gap, we introduce Grounding-Aware Sensitivity by Perturbation (GASP), a span-level detector that scores each answer sentence by how strongly its likelihood depends on the retrieved evidence, a quantity we term grounding sensitivity. GASP holds the answer fixed and re-scores it under the full context, under no context, and with each chunk removed, then measures the log-likelihood drops and Jensen-Shannon divergences (JSD). The likelihood of a grounded sentence collapses once its supporting passage is removed, whereas a hallucinated sentence is almost unaffected, a contrast we interpret by casting decoding as a random nonlinear iterated function system (RNIFS). We evaluate GASP on three benchmarks (RAGTruth, TofuEval, RAGBench) with three instruction-tuned scorers from two model families (Qwen2.5-0.5B, Qwen2.5-1.5B, and SmolLM2-1.7B) under a leakage-clean protocol. On RAGTruth it reaches a response-level area under the ROC curve (AUC) of about 0.73 and a span-level AUC of about 0.67, improving significantly over perplexity and by clear margins over length, whole-context natural language inference (NLI), and self-consistency baselines. The only baseline competitive at the span level is a well-configured chunk-level entailment verifier, which requires a separate model, whereas a training-free threshold on the grounding features matches the trained classifier without labeled data and serves as the default detector. Beyond RAGTruth, the signal transfers to TofuEval but not to short-answer question answering in RAGBench, showing GASP is best suited to outputs constructed from the retrieved context rather than answers recoverable from parametric knowledge.

Related papers

• When Explanations Betray Backdoors: Black-Box Auditing for Language Model Classifiers [4.106593007943804]
  We introduce Groundedness Drift, a lightweight score measuring whether the answer summary remains grounded in the input.<n>We then evaluate Unsupported Groundedness, a multi-probe escalation for explanation-camouflage stress cases.
  arXiv  Detail & Related papers  (2026-08-12T22:11:47Z)
• Latent Reward Registers for Diffusion Preference Alignment [46.49072106092868]
  We propose Latent Reward Registers, a mechanism that estimates terminal preference directly from intermediate noisy latents.<n>This independent readout mechanism extracts latent reward evidence without altering the generator's hidden states or velocity field.<n>For training, Reward-Gradient On-Policy Distillation (RG-OPD) distills reward-guided updates along on-policy trajectories, bypassing the computationally expensive rollouts of standard policy gradients.<n>For inference, Reward-Guided Sampling (RGS) steers trajectories via magnitude-matched reward gradients without parameter updates.
  arXiv  Detail & Related papers  (2026-08-04T17:00:52Z)
• What Accuracy and Gradient Cosine Miss: Evaluating Feedback Alignment via Scale Stability, Reference Validity, and Depth Utility [48.10132234701036]
  We propose a diagnostic evaluation protocol based on three checks -- scale stability, reference validity, and depth utility -- together with per-layer rather than aggregate cosine reporting.<n>Across multiple architectures and methods, our protocol identifies all failures with wide calibration margins.
  arXiv  Detail & Related papers  (2026-06-19T06:04:17Z)
• Retromorphic Testing with Hierarchical Verification for Hallucination Detection in RAG [3.2327627739464635]
  We present RT4CHART, a retromorphic testing framework for context-faithfulness assessment.<n>It decomposes model outputs into independently verifiable claims and performs hierarchical, local-to-global verification.<n>It achieves the best answer-level hallucination detection F1 among all baselines.
  arXiv  Detail & Related papers  (2026-03-29T16:12:18Z)
• ProGRank: Probe-Gradient Reranking to Defend Dense-Retriever RAG from Corpus Poisoning [5.725082598014574]
  Retrieval-Augmented Generation (RAG) improves reliability of large language model applications by grounding generation in retrieved evidence.<n>An adversary injects or edits passages so that they are ranked into the Top-$K$ results for target queries and then affect downstream generation.<n>We propose ProGRank, a post hoc, training-free retriever-side defence for dense-retriever RAG.
  arXiv  Detail & Related papers  (2026-03-24T08:29:15Z)
• Aligning to Illusions: Choice Blindness in Human and AI Feedback [2.912535004465788]
  In a human choice blindness study, 91% of surreptitiously swapped preferences go undetected, extending choice blindness to third-person evaluative comparison of unfamiliar text.<n>We find detection relies on shallow text matching rather than genuine self-monitoring.<n>In a dose-response experiment across two architectures, one-sixth to one-third of labels must be corrupted before the reward signal halves.<n>A Best-of-N evaluation confirms this translates to downstream policy degradation.
  arXiv  Detail & Related papers  (2026-03-09T14:10:36Z)
• Rethinking the Reranker: Boundary-Aware Evidence Selection for Robust Retrieval-Augmented Generation [64.09110141948693]
  Retrieval-Augmented Generation (RAG) systems remain brittle under realistic retrieval noise.<n>We propose BAR-RAG, which reframes the reranker as a boundary-aware evidence selector that targets the generator's Goldilocks Zone.<n>Bar-RAG consistently improves end-to-end performance under noisy retrieval.
  arXiv  Detail & Related papers  (2026-02-03T16:08:23Z)
• CARE What Fails: Contrastive Anchored-REflection for Verifiable Multimodal [84.71254539482369]
  Group-relative reinforcement learning with verifiable rewards (RLVR) often wastes the most informative data it already has the failures.<n>We present CARE, a failure-centric post-training framework for multimodal reasoning that turns errors into supervision.<n> CARE improves accuracy and training smoothness while explicitly increasing the share of learning signal that comes from failures.
  arXiv  Detail & Related papers  (2025-12-22T16:34:21Z)
• Fairness Testing in Retrieval-Augmented Generation: How Small Perturbations Reveal Bias in Small Language Models [0.1876920697241348]
  This study conducts fairness testing through metamorphic testing (MT), introducing controlled demographic perturbations in prompts to assess fairness in sentiment analysis performed by three Small Language Models (SLMs)<n>Results show that minor demographic variations can break up to one third of metamorphic relations (MRs)<n>A detailed analysis of these failures reveals a consistent bias hierarchy, with perturbations involving racial cues being the predominant cause of the violations.
  arXiv  Detail & Related papers  (2025-09-30T17:42:35Z)
• HyPoradise: An Open Baseline for Generative Speech Recognition with Large Language Models [81.56455625624041]
  We introduce the first open-source benchmark to utilize external large language models (LLMs) for ASR error correction. The proposed benchmark contains a novel dataset, HyPoradise (HP), encompassing more than 334,000 pairs of N-best hypotheses. LLMs with reasonable prompt and its generative capability can even correct those tokens that are missing in N-best list.
  arXiv  Detail & Related papers  (2023-09-27T14:44:10Z)
• The KFIoU Loss for Rotated Object Detection [115.334070064346]
  In this paper, we argue that one effective alternative is to devise an approximate loss who can achieve trend-level alignment with SkewIoU loss. Specifically, we model the objects as Gaussian distribution and adopt Kalman filter to inherently mimic the mechanism of SkewIoU. The resulting new loss called KFIoU is easier to implement and works better compared with exact SkewIoU.
  arXiv  Detail & Related papers  (2022-01-29T10:54:57Z)
• Constructing interval variables via faceted Rasch measurement and multitask deep learning: a hate speech application [63.10266319378212]
  We propose a method for measuring complex variables on a continuous, interval spectrum by combining supervised deep learning with the Constructing Measures approach to faceted Rasch item response theory (IRT) We demonstrate this new method on a dataset of 50,000 social media comments sourced from YouTube, Twitter, and Reddit and labeled by 11,000 U.S.-based Amazon Mechanical Turk workers.
  arXiv  Detail & Related papers  (2020-09-22T02:15:05Z)
• Stratified Rule-Aware Network for Abstract Visual Reasoning [46.015682319351676]
  Raven's Progressive Matrices (RPM) test is typically used to examine the capability of abstract reasoning. Recent studies, taking advantage of Convolutional Neural Networks (CNNs), have achieved encouraging progress to accomplish the RPM test. We propose a Stratified Rule-Aware Network (SRAN) to generate the rule embeddings for two input sequences.
  arXiv  Detail & Related papers  (2020-02-17T08:44:05Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.