From 4f52bdcd1b33a738a555e837043a17da0b3e3c97 Mon Sep 17 00:00:00 2001 From: anthology-assist Date: Wed, 18 Sep 2024 14:17:39 -0500 Subject: [PATCH] Paper Revision{2024.acl-long.329}, closes #3896. --- data/xml/2024.acl.xml | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/data/xml/2024.acl.xml b/data/xml/2024.acl.xml index b754558313..5237618275 100644 --- a/data/xml/2024.acl.xml +++ b/data/xml/2024.acl.xml @@ -4266,8 +4266,10 @@ AnetteFrankRuprecht-Karls-Universität Heidelberg 6048-6089 Large language models (LLMs) can explain their predictions through post-hoc or Chain-of-Thought (CoT) explanations. But an LLM could make up reasonably sounding explanations that are unfaithful to its underlying reasoning. Recent work has designed tests that aim to judge the faithfulness of post-hoc or CoT explanations. In this work we argue that these faithfulness tests do not measure faithfulness to the models’ inner workings – but rather their self-consistency at output level.Our contributions are three-fold: i) We clarify the status of faithfulness tests in view of model explainability, characterising them as self-consistency tests instead. This assessment we underline by ii) constructing a Comparative Consistency Bank for self-consistency tests that for the first time compares existing tests on a common suite of 11 open LLMs and 5 tasks – including iii) our new self-consistency measure CC-SHAP. CC-SHAP is a fine-grained measure (not a test) of LLM self-consistency. It compares how a model’s input contributes to the predicted answer and to generating the explanation. Our fine-grained CC-SHAP metric allows us iii) to compare LLM behaviour when making predictions and to analyse the effect of other consistency tests at a deeper level, which takes us one step further towards measuring faithfulness by bringing us closer to the internals of the model than strictly surface output-oriented tests. - 2024.acl-long.329 + 2024.acl-long.329 parcalabescu-frank-2024-measuring + + This revision mentions a sponsor in the acknowledgements and fixes the typo in Eq. 4. Learning or Self-aligning? Rethinking Instruction Fine-tuning