Consolidate LLM attr logic

pytorch / captum

Model interpretability and understanding for PyTorch

https://captum.ai

BSD 3-Clause "New" or "Revised" License

4.94k stars 499 forks source link

Closed craymichael closed 3 weeks ago

craymichael commented 3 weeks ago

Summary: Add base class BaseLLMAttribution to consolidate repeat logic between perturbation/gradient-based LLM attr classes

Differential Revision: D65008854

facebook-github-bot commented 3 weeks ago

This pull request was exported from Phabricator. Differential Revision: D65008854

facebook-github-bot commented 3 weeks ago

This pull request was exported from Phabricator. Differential Revision: D65008854

facebook-github-bot commented 3 weeks ago

This pull request has been merged in pytorch/captum@638b920f9cba6a286ea96f80383ddc166d341ae2.