Experience-Calibrated Contrastive Decoding for Mitigating Hallucinations in LM-Based Text-to-Speech
Language model-based text-to-speech (LM-based TTS) remains vulnerable to speech hallucinations that deviate from the target text. Existing mitigation mainly relies on architectural changes or additional training, while decoding-time control remains underexplored. We present a conditional information view that distinguishes text-derived alignment information from experience information supplied by acoustic context and learned speech regularities. We hypothesize that an important class of hallucin
Record details
Published: 1 August 2026
Source: arXiv cs.LG
Category: Research
Topics: Safety & alignment
Retrieved: 4 August 2026
Related evidence
These records share source-supplied organisations, an exact publisher byline, automatic topics or regions. The reason is shown on every link; related does not mean supporting, agreeing with or verifying this record.
AI Alignment and Fiduciary Obligation
arXiv alignment query · 1 August 2026
OpenART: Scaling Agent Red Teaming via Open-Ended Environment Evolution
arXiv red teaming query · 1 August 2026
FATE: Frame-Level Audio-Visual Temporal Embedding
HuggingFace Daily Papers · 1 August 2026
Alignment between innovation systems and entrepreneurial ecosystems in translating academic knowledge into entrepreneurial activity: Governing the knowledge filter
Technological Forecasting and Social Change · 1 August 2026
xMICD: Explainable Representation of Multiple ICD Codes
arXiv cs.LG · 2 August 2026
Mind the Gap: Zero-Query Jailbreaks via Filter-Generator Discrepancy in Text-to-Image Systems
arXiv red teaming query · 2 August 2026
How to cite this record
ethics.ai (1 August 2026), “Experience-Calibrated Contrastive Decoding for Mitigating Hallucinations in LM-Based Text-to-Speech,” evidence record 16161, https://ethics.ai/record/16161 (originally published by arXiv cs.LG).
Use and limitations
This page is a stable index and citation surface for a source record. ethics.ai did not author the underlying report and has not independently verified every claim. Automatic topics may be imperfect. For consequential use, quote and cite the original publisher.