@inproceedings{barkan-etal-2024-improving,
title = "Improving {LLM} Attributions with Randomized Path-Integration",
author = "Barkan, Oren and
Elisha, Yehonatan and
Toib, Yonatan and
Weill, Jonathan and
Koenigstein, Noam",
editor = "Al-Onaizan, Yaser and
Bansal, Mohit and
Chen, Yun-Nung",
booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2024",
month = nov,
year = "2024",
address = "Miami, Florida, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2024.findings-emnlp.551",
pages = "9430--9446",
abstract = "We present Randomized Path-Integration (RPI) - a path-integration method for explaining language models via randomization of the integration path over the attention information in the model. RPI employs integration on internal attention scores and their gradients along a randomized path, which is dynamically established between a baseline representation and the attention scores of the model. The inherent randomness in the integration path originates from modeling the baseline representation as a randomly drawn tensor from a Gaussian diffusion process. As a consequence, RPI generates diverse baselines, yielding a set of candidate attribution maps. This set facilitates the selection of the most effective attribution map based on the specific metric at hand. We present an extensive evaluation, encompassing 11 explanation methods and 5 language models, including the Llama2 and Mistral models. Our results demonstrate that RPI outperforms latest state-of-the-art methods across 4 datasets and 5 evaluation metrics.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="barkan-etal-2024-improving">
<titleInfo>
<title>Improving LLM Attributions with Randomized Path-Integration</title>
</titleInfo>
<name type="personal">
<namePart type="given">Oren</namePart>
<namePart type="family">Barkan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yehonatan</namePart>
<namePart type="family">Elisha</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yonatan</namePart>
<namePart type="family">Toib</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jonathan</namePart>
<namePart type="family">Weill</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Noam</namePart>
<namePart type="family">Koenigstein</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2024-11</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2024</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yaser</namePart>
<namePart type="family">Al-Onaizan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohit</namePart>
<namePart type="family">Bansal</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yun-Nung</namePart>
<namePart type="family">Chen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Miami, Florida, USA</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>We present Randomized Path-Integration (RPI) - a path-integration method for explaining language models via randomization of the integration path over the attention information in the model. RPI employs integration on internal attention scores and their gradients along a randomized path, which is dynamically established between a baseline representation and the attention scores of the model. The inherent randomness in the integration path originates from modeling the baseline representation as a randomly drawn tensor from a Gaussian diffusion process. As a consequence, RPI generates diverse baselines, yielding a set of candidate attribution maps. This set facilitates the selection of the most effective attribution map based on the specific metric at hand. We present an extensive evaluation, encompassing 11 explanation methods and 5 language models, including the Llama2 and Mistral models. Our results demonstrate that RPI outperforms latest state-of-the-art methods across 4 datasets and 5 evaluation metrics.</abstract>
<identifier type="citekey">barkan-etal-2024-improving</identifier>
<location>
<url>https://aclanthology.org/2024.findings-emnlp.551</url>
</location>
<part>
<date>2024-11</date>
<extent unit="page">
<start>9430</start>
<end>9446</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Improving LLM Attributions with Randomized Path-Integration
%A Barkan, Oren
%A Elisha, Yehonatan
%A Toib, Yonatan
%A Weill, Jonathan
%A Koenigstein, Noam
%Y Al-Onaizan, Yaser
%Y Bansal, Mohit
%Y Chen, Yun-Nung
%S Findings of the Association for Computational Linguistics: EMNLP 2024
%D 2024
%8 November
%I Association for Computational Linguistics
%C Miami, Florida, USA
%F barkan-etal-2024-improving
%X We present Randomized Path-Integration (RPI) - a path-integration method for explaining language models via randomization of the integration path over the attention information in the model. RPI employs integration on internal attention scores and their gradients along a randomized path, which is dynamically established between a baseline representation and the attention scores of the model. The inherent randomness in the integration path originates from modeling the baseline representation as a randomly drawn tensor from a Gaussian diffusion process. As a consequence, RPI generates diverse baselines, yielding a set of candidate attribution maps. This set facilitates the selection of the most effective attribution map based on the specific metric at hand. We present an extensive evaluation, encompassing 11 explanation methods and 5 language models, including the Llama2 and Mistral models. Our results demonstrate that RPI outperforms latest state-of-the-art methods across 4 datasets and 5 evaluation metrics.
%U https://aclanthology.org/2024.findings-emnlp.551
%P 9430-9446
Markdown (Informal)
[Improving LLM Attributions with Randomized Path-Integration](https://aclanthology.org/2024.findings-emnlp.551) (Barkan et al., Findings 2024)
ACL
- Oren Barkan, Yehonatan Elisha, Yonatan Toib, Jonathan Weill, and Noam Koenigstein. 2024. Improving LLM Attributions with Randomized Path-Integration. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 9430–9446, Miami, Florida, USA. Association for Computational Linguistics.