Matches in SemOpenAlex for { <https://semopenalex.org/work/W4387075008> ?p ?o ?g. }
Showing items 1 to 59 of
59
with 100 items per page.
- W4387075008 abstract "Large language models (LLMs) are increasingly being used for tasks beyond text generation, including complex tasks such as data labeling, information extraction, etc. With the recent surge in research efforts to comprehend the full extent of LLM capabilities, in this work, we investigate the role of LLMs as counterfactual explanation modules, to explain decisions of black-box text classifiers. Inspired by causal thinking, we propose a pipeline for using LLMs to generate post-hoc, model-agnostic counterfactual explanations in a principled way via (i) leveraging the textual understanding capabilities of the LLM to identify and extract latent features, and (ii) leveraging the perturbation and generation capabilities of the same LLM to generate a counterfactual explanation by perturbing input features derived from the extracted latent features. We evaluate three variants of our framework, with varying degrees of specificity, on a suite of state-of-the-art LLMs, including ChatGPT and LLaMA 2. We evaluate the effectiveness and quality of the generated counterfactual explanations, over a variety of text classification benchmarks. Our results show varied performance of these models in different settings, with a full two-step feature extraction based variant outperforming others in most cases. Our pipeline can be used in automated explanation systems, potentially reducing human effort." @default.
- W4387075008 created "2023-09-27" @default.
- W4387075008 creator A5013881064 @default.
- W4387075008 creator A5035107313 @default.
- W4387075008 creator A5047078657 @default.
- W4387075008 creator A5086378853 @default.
- W4387075008 date "2023-09-23" @default.
- W4387075008 modified "2023-09-28" @default.
- W4387075008 title "LLMs as Counterfactual Explanation Modules: Can ChatGPT Explain Black-box Text Classifiers?" @default.
- W4387075008 doi "https://doi.org/10.48550/arxiv.2309.13340" @default.
- W4387075008 hasPublicationYear "2023" @default.
- W4387075008 type Work @default.
- W4387075008 citedByCount "0" @default.
- W4387075008 crossrefType "posted-content" @default.
- W4387075008 hasAuthorship W4387075008A5013881064 @default.
- W4387075008 hasAuthorship W4387075008A5035107313 @default.
- W4387075008 hasAuthorship W4387075008A5047078657 @default.
- W4387075008 hasAuthorship W4387075008A5086378853 @default.
- W4387075008 hasBestOaLocation W43870750081 @default.
- W4387075008 hasConcept C108650721 @default.
- W4387075008 hasConcept C119857082 @default.
- W4387075008 hasConcept C154945302 @default.
- W4387075008 hasConcept C15744967 @default.
- W4387075008 hasConcept C180747234 @default.
- W4387075008 hasConcept C199360897 @default.
- W4387075008 hasConcept C204321447 @default.
- W4387075008 hasConcept C2522767166 @default.
- W4387075008 hasConcept C41008148 @default.
- W4387075008 hasConcept C43521106 @default.
- W4387075008 hasConcept C77805123 @default.
- W4387075008 hasConcept C94966114 @default.
- W4387075008 hasConceptScore W4387075008C108650721 @default.
- W4387075008 hasConceptScore W4387075008C119857082 @default.
- W4387075008 hasConceptScore W4387075008C154945302 @default.
- W4387075008 hasConceptScore W4387075008C15744967 @default.
- W4387075008 hasConceptScore W4387075008C180747234 @default.
- W4387075008 hasConceptScore W4387075008C199360897 @default.
- W4387075008 hasConceptScore W4387075008C204321447 @default.
- W4387075008 hasConceptScore W4387075008C2522767166 @default.
- W4387075008 hasConceptScore W4387075008C41008148 @default.
- W4387075008 hasConceptScore W4387075008C43521106 @default.
- W4387075008 hasConceptScore W4387075008C77805123 @default.
- W4387075008 hasConceptScore W4387075008C94966114 @default.
- W4387075008 hasLocation W43870750081 @default.
- W4387075008 hasOpenAccess W4387075008 @default.
- W4387075008 hasPrimaryLocation W43870750081 @default.
- W4387075008 hasRelatedWork W2020540721 @default.
- W4387075008 hasRelatedWork W2961085424 @default.
- W4387075008 hasRelatedWork W2992516105 @default.
- W4387075008 hasRelatedWork W3207353404 @default.
- W4387075008 hasRelatedWork W4285260836 @default.
- W4387075008 hasRelatedWork W4285281025 @default.
- W4387075008 hasRelatedWork W4286629047 @default.
- W4387075008 hasRelatedWork W4306321456 @default.
- W4387075008 hasRelatedWork W4306674287 @default.
- W4387075008 hasRelatedWork W4224009465 @default.
- W4387075008 isParatext "false" @default.
- W4387075008 isRetracted "false" @default.
- W4387075008 workType "article" @default.