Balepur, Nishant; Ravichander, Abhilasha; Rudinger, Rachel
Artifacts or Abduction: How Do LLMs Answer Multiple-Choice Questions Without the Question? Sonstige
In-progress preprint, 2024.
Abstract | Links | BibTeX | Schlagwörter: artificial intelligence, KI, large language models, LLM, multiple choice, O
@misc{balepur2024artifacts,
title = {Artifacts or Abduction: How Do LLMs Answer Multiple-Choice Questions Without the Question?},
author = {Nishant Balepur and Abhilasha Ravichander and Rachel Rudinger},
url = {https://doi.org/10.48550/arXiv.2402.12483},
doi = {10.48550/arXiv.2402.12483 Focus to learn more},
year = {2024},
date = {2024-01-01},
urldate = {2024-01-01},
abstract = {Multiple-choice question answering (MCQA) is often used to evaluate large language models (LLMs). To see if MCQA assesses LLMs as intended, we probe if LLMs can perform MCQA with choices-only prompts, where models must select the correct answer only from the choices. In three MCQA datasets and four LLMs, this prompt bests a majority baseline in 11/12 cases, with up to 0.33 accuracy gain. To help explain this behavior, we conduct an in-depth, black-box analysis on memorization, choice dynamics, and question inference. Our key findings are threefold. First, we find no evidence that the choices-only accuracy stems from memorization alone. Second, priors over individual choices do not fully explain choices-only accuracy, hinting that LLMs use the group dynamics of choices. Third, LLMs have some ability to infer a relevant question from choices, and surprisingly can sometimes even match the original question. We hope to motivate the use of stronger baselines in MCQA benchmarks, the design of robust MCQA datasets, and further efforts to explain LLM decision-making.},
howpublished = {In-progress preprint},
keywords = {artificial intelligence, KI, large language models, LLM, multiple choice, O},
pubstate = {published},
tppubtype = {misc}
}
Limburg, Anika; Salden, Peter; Mundorf, Margret; Weßels, Doris
Plagiarismus in Zeiten Künstlicher Intelligenz Artikel
In: Zeitschrift für Hochschulentwicklung (ZFHE), Bd. 17, Ausg. 3, S. 91–106, 2022, ISSN: 2219-6994.
Abstract | Links | BibTeX | Schlagwörter: A, AI, artificial intelligence, gute wissenschaftliche Praxis, KI, Künstliche Intelligenz, Natural Language Processing, plagiarism, Plagiarismus, reflective science, Schreibdidaktik, writing tools
@article{Limburg2022,
title = {Plagiarismus in Zeiten Künstlicher Intelligenz},
author = {Anika Limburg and Peter Salden and Margret Mundorf and Doris Weßels},
editor = {Ines Langemeyer and Ernst Schraube and Peter Tremp},
url = {https://doi.org/10.3217/zfhe-17-03/06},
doi = {10.3217/zfhe-17-03/06},
issn = {2219-6994},
year = {2022},
date = {2022-10-01},
journal = {Zeitschrift für Hochschulentwicklung (ZFHE)},
volume = {17},
issue = {3},
pages = {91–106},
abstract = {Software auf Basis Künstlicher Intelligenz aus dem Bereich des Natural Language Processing hat das Potenzial, wissenschaftliches Schreiben grundlegend zu verändern. Entsprechende Tools können bereits erstaunlich kohärente Texte in wissenschaftlichem Ton produzieren. Dies führt zu fundamentalen Fragen guter wissenschaftlicher Praxis und akademischer Kultur. Wir diskutieren diese Entwicklung vor dem Hintergrund einer Befragung deutscher Schreibdidaktiker:innen und arbeiten Fragen heraus, die im Zusammenhang mit KI-Schreibtools zukünftig von zentraler Bedeutung sein werden. Abschließend schlagen wir einen Passus für eine Selbstständigkeitserklärung vor, der den Entwicklungen Rechnung trägt.},
keywords = {A, AI, artificial intelligence, gute wissenschaftliche Praxis, KI, Künstliche Intelligenz, Natural Language Processing, plagiarism, Plagiarismus, reflective science, Schreibdidaktik, writing tools},
pubstate = {published},
tppubtype = {article}
}