EvalTree
An advanced taxonomy construction method that builds a classification tree by hierarchically clustering user prompts and using an LLM to generate a name for each cluster, with cluster names serving as node labels.
Timeline (last 90 days)
On the WildBench benchmark, COMPREHEND's entailment/faithfulness metric rises from 22.6 for EvalTree to 41.7.
EvalTree is an advanced taxonomy-construction method that builds a taxonomy tree by hierarchically clustering user prompts and using an LLM to generate a name for each cluster.
When predicting response scores in the benchmark tests, the weakness reports generated by COMPREHEND are more informative than those of EvalTree.
All Facts (3)
On the WildBench benchmark, COMPREHEND's entailment/faithfulness metric rises from 22.6 for EvalTree to 41.7.
50%UnverifiedEvalTree is an advanced taxonomy-construction method that builds a taxonomy tree by hierarchically clustering user prompts and using an LLM to generate a name for each cluster.
50%UnverifiedWhen predicting response scores in the benchmark tests, the weakness reports generated by COMPREHEND are more informative than those of EvalTree.
50%