TDGNet: Hallucination Detection in Diffusion Language Models via Temporal Dynamic Graphs

• URL: http://arxiv.org/abs/2602.08048v1
• Date: Sun, 08 Feb 2026 16:35:30 GMT
• Title: TDGNet: Hallucination Detection in Diffusion Language Models via Temporal Dynamic Graphs
• Authors: Arshia Hemmat, Philip Torr, Yongqiang Chen, Junchi Yu,
• Abstract summary: Diffusion language models (D-LLMs) offer parallel denoising and bidirectional context.<n> hallucination detection for D-LLMs remains underexplored.<n>We introduce TDGNet, a temporal dynamic graph framework that formulates hallucination detection as learning over evolving token-level attention graphs.
• Score: 30.313604786976715
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Diffusion language models (D-LLMs) offer parallel denoising and bidirectional context, but hallucination detection for D-LLMs remains underexplored. Prior detectors developed for auto-regressive LLMs typically rely on single-pass cues and do not directly transfer to diffusion generation, where factuality evidence is distributed across the denoising trajectory and may appear, drift, or be self-corrected over time. We introduce TDGNet, a temporal dynamic graph framework that formulates hallucination detection as learning over evolving token-level attention graphs. At each denoising step, we sparsify the attention graph and update per-token memories via message passing, then apply temporal attention to aggregate trajectory-wide evidence for final prediction. Experiments on LLaDA-8B and Dream-7B across QA benchmarks show consistent AUROC improvements over output-based, latent-based, and static-graph baselines, with single-pass inference and modest overhead. These results highlight the importance of temporal reasoning on attention graphs for robust hallucination detection in diffusion language models.

Related papers

• Sequence Diffusion Model for Temporal Link Prediction in Continuous-Time Dynamic Graph [5.83093727437226]
  Existing temporal graph neural networks mainly focus on learning representations of historical interactions.<n>We propose a novel sequence-level diffusion framework that unifies dynamic graph learning with generative denoising.<n>We show that our framework consistently achieves state-of-the-art performance in the temporal link prediction task.
  arXiv  Detail & Related papers  (2026-01-30T18:02:12Z)
• Hallucination Begins Where Saliency Drops [18.189047289404325]
  hallucinations frequently arise when preceding output tokens exhibit low saliency toward the prediction of the next token.<n>We introduce LVLMs-Saliency, a gradient-aware diagnostic framework that quantifies the visual grounding strength of each output token.<n>Our method significantly reduces hallucination rates while preserving fluency and task performance, offering a robust and interpretable solution.
  arXiv  Detail & Related papers  (2026-01-28T05:50:52Z)
• TAG:Tangential Amplifying Guidance for Hallucination-Resistant Diffusion Sampling [53.61290359948953]
  Tangential Amplifying Guidance (TAG) operates solely on trajectory signals without modifying the underlying diffusion model.<n>We formalize this guidance process by leveraging a first-order Taylor expansion.<n> TAG is a plug-and-play, architecture-agnostic module that improves diffusion sampling fidelity with minimal computational addition.
  arXiv  Detail & Related papers  (2025-10-06T06:53:29Z)
• TraceDet: Hallucination Detection from the Decoding Trace of Diffusion Large Language Models [49.83690850047884]
  hallucination problem in D-LLMs remains underexplored, limiting their reliability in real-world applications.<n>Existing hallucination detection methods are designed for AR-LLMs and rely on signals from single-step generation.<n>We propose TraceDet, a novel framework that explicitly leverages the intermediate denoising steps of D-LLMs for hallucination detection.
  arXiv  Detail & Related papers  (2025-09-30T02:01:10Z)
• Neural Message-Passing on Attention Graphs for Hallucination Detection [32.29963721910821]
  CHARM casts hallucination detection as a graph learning task and tackles it by applying GNNs over the above attributed graphs.<n>We show that CHARM provably subsumes prior attention-based traces and, experimentally, it consistently outperforms other approaches across diverse benchmarks.
  arXiv  Detail & Related papers  (2025-09-29T13:37:12Z)
• EigenTrack: Spectral Activation Feature Tracking for Hallucination and Out-of-Distribution Detection in LLMs and VLMs [8.616813040714883]
  EigenTrack is an interpretable real-time detector for large language models (LLMs)<n>It tracks temporal shifts in representation structure that signal hallucination and OOD drift before surface errors appear.<n>Unlike existing white-box detectors, it preserves temporal context, aggregates global signals, and offers interpretable accuracy-latency trade-offs.
  arXiv  Detail & Related papers  (2025-09-19T08:05:28Z)
• Multimodal LLM-Guided Semantic Correction in Text-to-Image Diffusion [52.315729095824906]
  MLLM Semantic-Corrected Ping-Pong-Ahead Diffusion (PPAD) is a novel framework that introduces a Multimodal Large Language Model (MLLM) as a semantic observer during inference.<n>It performs real-time analysis on intermediate generations, identifies latent semantic inconsistencies, and translates feedback into controllable signals that actively guide the remaining denoising steps.<n>Extensive experiments demonstrate PPAD's significant improvements.
  arXiv  Detail & Related papers  (2025-05-26T14:42:35Z)
• DyG-Mamba: Continuous State Space Modeling on Dynamic Graphs [59.434893231950205]
  Dynamic graph learning aims to uncover evolutionary laws in real-world systems. We propose DyG-Mamba, a new continuous state space model for dynamic graph learning. We show that DyG-Mamba achieves state-of-the-art performance on most datasets.
  arXiv  Detail & Related papers  (2024-08-13T15:21:46Z)
• Detecting Anomalies in Dynamic Graphs via Memory enhanced Normality [39.476378833827184]
  Anomaly detection in dynamic graphs presents a significant challenge due to the temporal evolution of graph structures and attributes. We introduce a novel spatial- temporal memories-enhanced graph autoencoder (STRIPE) STRIPE significantly outperforms existing methods with 5.8% improvement in AUC scores and 4.62X faster in training time.
  arXiv  Detail & Related papers  (2024-03-14T02:26:10Z)
• Data Attribution for Diffusion Models: Timestep-induced Bias in Influence Estimation [53.27596811146316]
  Diffusion models operate over a sequence of timesteps instead of instantaneous input-output relationships in previous contexts. We present Diffusion-TracIn that incorporates this temporal dynamics and observe that samples' loss gradient norms are highly dependent on timestep. We introduce Diffusion-ReTrac as a re-normalized adaptation that enables the retrieval of training samples more targeted to the test sample of interest.
  arXiv  Detail & Related papers  (2024-01-17T07:58:18Z)
• Anticipating the Unseen Discrepancy for Vision and Language Navigation [63.399180481818405]
  Vision-Language Navigation requires the agent to follow natural language instructions to reach a specific target. The large discrepancy between seen and unseen environments makes it challenging for the agent to generalize well. We propose Unseen Discrepancy Anticipating Vision and Language Navigation (DAVIS) that learns to generalize to unseen environments via encouraging test-time visual consistency.
  arXiv  Detail & Related papers  (2022-09-10T19:04:40Z)

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.