Post
107
Zero-shot encoder models cannot provide information that elucidates why they chose their classification.
This is a first of it's kind model that addresses that issue directly. dleemiller/AbsoluteZero-nli-4b
AbsoluteZero-nli outputs traces that gives observability to NLI.
Traces enable users to inspect and determine how to augment hypotheses to fix or improve classification. This could be done in conjunction with automation for grouping classification failures and using an LLM to test changes to the hypothesis against a holdout set.
🥇https://huggingface.co/spaces/dleemiller/open-nli-leaderboard
This is a first of it's kind model that addresses that issue directly. dleemiller/AbsoluteZero-nli-4b
AbsoluteZero-nli outputs traces that gives observability to NLI.
Traces enable users to inspect and determine how to augment hypotheses to fix or improve classification. This could be done in conjunction with automation for grouping classification failures and using an LLM to test changes to the hypothesis against a holdout set.
🥇https://huggingface.co/spaces/dleemiller/open-nli-leaderboard