ClariEdge, ASR data selection and contextual PII masking show a shift toward selecting representative training subsets and contextual labels for specialized models. Evaluation frameworks also reveal that LLM judges can sharply change perceived tutoring success, making metrics a strategic choice.
The ReCoPII framework builds contextual PII masking datasets without the author-based augmentation premise. On Korean documents from six domains, sampled KEEP labels reach 88.42% annotator agreement, against 26.27% for SPY-style augmentation. Classif...
An embedding-based method using Whisper's encoder selects smaller, representative training subsets for ASR fine-tuning. In Hebrew and Danish, k-center selection often outperforms random sampling, especially at smaller budgets. Lexical diversity analy...
A fully automated framework shows that on 5,000+ simulated dialogues, Remediation Success Rate moves from 42.1% under exact-match to 81.3% under human-calibrated LLM judges. The 39.2-point gap exceeds the combined effect of model, teaching strategy, ...
IBM Research releases ClariEdge, a dataset of about 3,900 examples for spotting ambiguous queries in RAG. A lightweight LoRA obtained through fine-tuning reaches 96.6% accuracy and outperforms larger baselines. The signal: specialized local component...
IBM Research has released ClariEdge, a dataset of about 3,900 examples for recognizing ambiguous queries in RAG systems and generating clarification requests. A lightweight LoRA reaches 96.6% classification accuracy and beats prompted baselines from ...
