VLP adds an NL documentation layer with trace-linked mismatch detection and derived formal checks to make human validation of LLM code feasible, lifting pass@1 from 28.7-73.2% to 65.4-93.5%.
Fewer Hallucinations, More Verification: A Three-Stage LLM-Based Framework for ASR Error Correction,
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
years
2026 3verdicts
UNVERDICTED 3representative citing papers
Internal decoder probing of Whisper yields strongest hallucination detection without references, with late fusion of text and internal features performing best overall.
G-SPIN uses a GNN to restrict ASR correction search to phonetic neighbors, then applies MLM local scoring and LLM re-ranking for context-aware fixes at inference time.
citing papers explorer
-
Guiding Human Validation of LLM-Generated Code via Verifiable Literate Programming
VLP adds an NL documentation layer with trace-linked mismatch detection and derived formal checks to make human validation of LLM code feasible, lifting pass@1 from 28.7-73.2% to 65.4-93.5%.
-
From Text Metrics to Model Internals: A Study of Whisper ASR Hallucination Detection
Internal decoder probing of Whisper yields strongest hallucination detection without references, with late fusion of text and internal features performing best overall.
-
Graph-Based Phonetic Error Correction of Noisy ASR
G-SPIN uses a GNN to restrict ASR correction search to phonetic neighbors, then applies MLM local scoring and LLM re-ranking for context-aware fixes at inference time.