Weakly Supervised Distillation of Hallucination Signals into Transformer Representations

Computer Science > Artificial Intelligence [Submitted on 7 Apr 2026] Weakly Supervised Distillation of Hallucination Signals into Transformer Representations Shoaib Sadiq Salehmohamed, Jinal Prashant Thakkar, Hansika Aredla, Shaik Mohammed Omar, Shalmali Ayachit Existing hallucination detection methods for large language models (LLMs) rely on external verification at inference time, requiring gold answers, retrieval systems, or auxiliary judge models. We ask whether this external supervision can instead be distilled into the model's own representations during training, enabling hallucination detection from internal activations alone at inference time. We introduce a weak supervision framework that combines three complementary grounding signals: substring matching, sentence embedding similarity, and an LLM as a judge verdict to label generated responses as grounded or hallucinated without human annotation. Using this framework, we construct a 15000-sample dataset from SQuAD v2 (10500 train/development samples and a separate 5000-sample test set), where each example pairs a LLaMA-2-7B generated answer with its full per-layer hidden states and structured hallucination labels. We then train five probing classifiers: ProbeMLP (M0), LayerWiseMLP (M1), CrossLayerTransformer (M2), HierarchicalTransformer (M3), and CrossLayerAttentionTransformerV2 (M4), directly on these hidden states, treating external grounding signals as training-time supervision only. Our central hypothesis is that hallucination detection signals can be distilled into transformer representations, enabling internal detection without any external verification at inference time. Results support this hypothesis. Transformer-based probes achieve the strongest discrimination, with M2 performing best on 5-fold average AUC/F1, and M3 performing best on both single-fold validation and held-out test evaluation. We also benchmark inference efficiency: probe latency ranges from 0.15 to 5.62 ms (batched) and 1.55 to 6.66 ms (single sample), while end-to-end generation plus probe throughput remains approximately 0.231 queries per second, indicating negligible practical overhead. Comments: 20 pages, 6 figures, 6 tables. Introduces a 15k-sample representation-level hallucination dataset with full transformer hidden states and multi-signal weak supervision. Evaluates 5 probing architectures and demonstrates internal hallucination detection without external inference-time signals. Includes held-out test evaluation and deployment benchmarks Subjects: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG) ACM classes: I.2.6; I.2.7 Cite as: arXiv:2604.06277 [cs.AI]   (or arXiv:2604.06277v1 [cs.AI] for this version)   https://doi.org/10.48550/arXiv.2604.06277 Focus to learn more Submission history From: Shoaib Sadiq Salehmohamed [view email] [v1] Tue, 7 Apr 2026 08:14:48 UTC (28,449 KB) Access Paper: HTML (experimental) view license Current browse context: cs.AI < prev   |   next > new | recent | 2026-04 Change to browse by: cs cs.CL cs.LG References & Citations NASA ADS Google Scholar Semantic Scholar Export BibTeX Citation Bookmark Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer (What is the Explorer?) Connected Papers Toggle Connected Papers (What is Connected Papers?) Litmaps Toggle Litmaps (What is Litmaps?) scite.ai Toggle scite Smart Citations (What are Smart Citations?) Code, Data, Media Demos Related Papers About arXivLabs Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)