1 results

temporal-consistency

inall models datasets apps agents people projects

Datasets

All datasets matching “temporal-consistency”

mih2034 /temporal-consistency Temporal Consistency Evaluation for Multi-Turn Dialogue (AXCEL + PairEval) This repository contains the datasets and experiment results for evaluating temporal consistency in multi-turn dialogue. The work compares two LLM-as-a-judge metrics: AXCEL (fact-based consistency scoring) PairEval (pairwise preference scoring) All experiments evaluate the last assistant message against the preceding dialogue history. What is included Synthetic multi-turn dialogue datasets… See the full description on the dataset page: https://huggingface.co/datasets/mih2034/temporal-consistency.textn<1K0 likes4 downloads7mo agoHugging Face