{"status":"ok","message-type":"work-list","message-version":"1.0.0","message":{"facets":{},"total-results":7011590,"items":[{"indexed":{"date-parts":[[2026,10,2]],"date-time":"2026-10-02T14:00:58Z","timestamp":1790949658407,"version":"4.1.0"},"reference-count":50,"publisher":"SAGE Publications","license":[{"start":{"date-parts":[[2026,9,29]],"date-time":"2026-09-29T00:00:00Z","timestamp":1790640000000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/journals.sagepub.com\/page\/policies\/text-and-data-mining-license"}],"content-domain":{"domain":["journals.sagepub.com"],"crossmark-restriction":true},"short-container-title":["Sociological Methods &amp; Research"],"abstract":"<jats:p>\n                    When a language model answers a sequence of questions, researchers report the model and temperature but rarely what or how it was told about its own earlier answers. I show that this underreported setting behaves as a mode effect. Testing four large language model (LLM) families with revealed preference theory (GARP) across moral, economic, and social tasks (\n                    <jats:inline-formula>\n                      <mml:math xmlns:mml=\"http:\/\/www.w3.org\/1998\/Math\/MathML\" display=\"inline\" overflow=\"scroll\">\n                        <mml:mi>N<\/mml:mi>\n                        <mml:mo>=<\/mml:mo>\n                        <mml:mn>5,926<\/mml:mn>\n                      <\/mml:math>\n                    <\/jats:inline-formula>\n                    ), I find that how prior choices re-enter the prompt, the\n                    <jats:italic toggle=\"yes\">memory condition<\/jats:italic>\n                    , moves the share of internally consistent subjects by up to 48 percentage points, mostly where that share starts lowest. Memory also influences which preferences the models express and, in a silicon-sampling extension, it brings the answers closer to the human distribution. Nonetheless, gains in consistency and surface fidelity do not necessarily translate into representativeness: a consistent, human-looking model may not represent any human population. For LLMs as social-science instruments, this study offers promise, caution, and a method for testing the stability of sequential outputs.\n                  <\/jats:p>","DOI":"10.1177\/00491241261492203","type":"journal-article","created":{"date-parts":[[2026,9,29]],"date-time":"2026-09-29T11:53:53Z","timestamp":1790682833000},"update-policy":"https:\/\/doi.org\/10.1177\/sage-journals-update-policy","source":"Crossref","is-referenced-by-count":0,"title":["How Model Choice and Memory Shape Preference Consistency in Large Language Models"],"prefix":"10.1177","author":[{"ORCID":"https:\/\/orcid.org\/0009-0007-9286-279X","authenticated-orcid":false,"given":"Yick","family":"Chung","sequence":"first","affiliation":[{"name":"University of Milan"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"179","published-online":{"date-parts":[[2026,9,29]]},"reference":[{"key":"e_1_3_4_2_1","doi-asserted-by":"publisher","DOI":"10.1177\/25152459251325174"},{"key":"e_1_3_4_3_1","doi-asserted-by":"publisher","DOI":"10.2307\/2525382"},{"key":"e_1_3_4_4_1","doi-asserted-by":"publisher","DOI":"10.1038\/s41562-025-02172-y"},{"key":"e_1_3_4_5_1","doi-asserted-by":"publisher","DOI":"10.1111\/1468-0262.00302"},{"key":"e_1_3_4_6_1","doi-asserted-by":"publisher","DOI":"10.1017\/pan.2023.2"},{"key":"e_1_3_4_7_1","unstructured":"Bai Yuntao Kadavath Saurav Kundu Sandipan Askell Amanda Kernion Jackson Jones Andy Chen Anna Goldie Anna Mirhoseini Azalia McKinnon Cameron. 2022. \u201cConstitutional AI: Harmlessness From AI Feedback.\u201d arXiv preprint arXiv:2212.08073."},{"key":"e_1_3_4_8_1","doi-asserted-by":"publisher","DOI":"10.1177\/00491241251330582"},{"key":"e_1_3_4_9_1","first-page":"1877","article-title":"Language Models are Few-Shot Learners","volume":"33","author":"Brown Tom B.","year":"2020","unstructured":"Brown Tom B., Benjamin Mann Nick Ryder, Melanie Subbiah Jared Kaplan, Prafulla Dhariwal Arvind Neelakantan, Pranav Shyam Girish Sastry, Askell Amanda. 2020. \u201cLanguage Models are Few-Shot Learners.\u201d Advances in Neural Information Processing Systems 33, 1877\u20131901.","journal-title":"Advances in Neural Information Processing Systems"},{"key":"e_1_3_4_10_1","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.2316205120"},{"key":"e_1_3_4_11_1","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.2412015122"},{"key":"e_1_3_4_12_1","doi-asserted-by":"publisher","DOI":"10.1257\/aer.104.6.1518"},{"key":"e_1_3_4_13_1","unstructured":"Chung Yick. 2026. \u201cReplication Package For: How Model Choice and Memory Shape Preference Consistency in Large Language Models.\u201d doi:10.5281\/zenodo.19004452."},{"key":"e_1_3_4_14_1","doi-asserted-by":"publisher","DOI":"10.1126\/science.adq1814"},{"key":"e_1_3_4_15_1","doi-asserted-by":"publisher","DOI":"10.1177\/00491241251339184"},{"key":"e_1_3_4_16_1","doi-asserted-by":"publisher","DOI":"10.1002\/9781394260645"},{"issue":"1","key":"e_1_3_4_17_1","first-page":"61","article-title":"Accurate Methods for the Statistics of Surprise and Coincidence","volume":"19","author":"Dunning Ted.","year":"1993","unstructured":"Dunning Ted. 1993. \u201cAccurate Methods for the Statistics of Surprise and Coincidence.\u201d Computational Linguistics 19(1): 61\u201374.","journal-title":"Computational Linguistics"},{"key":"e_1_3_4_18_1","doi-asserted-by":"crossref","unstructured":"Fisman Raymond Kariv Shachar Markovits Daniel. 2007b. \u201cReplication Data For: Individual Preferences for Giving.\u201d doi:10.3886\/E116294V1.","DOI":"10.1257\/aer.97.5.1858"},{"key":"e_1_3_4_19_1","doi-asserted-by":"publisher","DOI":"10.1257\/aer.97.5.1858"},{"key":"e_1_3_4_20_1","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.2501660122"},{"key":"e_1_3_4_21_1","doi-asserted-by":"publisher","DOI":"10.1214\/08-AOAS191"},{"key":"e_1_3_4_22_1","doi-asserted-by":"publisher","DOI":"10.1126\/sciadv.adz2924"},{"key":"e_1_3_4_23_1","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.2305016120"},{"key":"e_1_3_4_24_1","doi-asserted-by":"publisher","DOI":"10.1073\/pnas.2214005119"},{"key":"e_1_3_4_25_1","unstructured":"Guzm\u00e1n Ricardo A. Cosmides Leda. 2022. \u201cMoral Tradeoff System: Data and Scripts for Moral Tradeoff System Paper in PNAS.\u201d https:\/\/osf.io\/kd34j\/."},{"key":"e_1_3_4_26_1","doi-asserted-by":"publisher","DOI":"10.1038\/s41586-025-10021-1"},{"key":"e_1_3_4_27_1","unstructured":"Haerpfer Christian Inglehart Ronald Moreno Alejandro Welzel Christian Kizilova Kseniya Diez-Medrano Jaime Lagos Marta Norris Pippa Ponarin Eduard Puranen Bi. 2022. \u201cWorld Values Survey Wave 7 (2017\u20132022) Cross-National Data-Set.\u201d doi:10.14281\/18241.20."},{"key":"e_1_3_4_28_1","doi-asserted-by":"publisher","DOI":"10.1017\/S0140525X0999152X"},{"key":"e_1_3_4_29_1","doi-asserted-by":"crossref","unstructured":"Horton John J. Filippas Apostolos Manning Benjamin S.. 2023. \u201cLarge Language Models as Simulated Economic Agents: What Can We Learn From Homo Silicus?.\u201d Working Paper 31122 National Bureau of Economic Research. doi:10.3386\/w31122.","DOI":"10.3386\/w31122"},{"key":"e_1_3_4_30_1","volume-title":"Theory of Probability","author":"Jeffreys Harold.","year":"1961","unstructured":"Jeffreys Harold. 1961. Theory of Probability. 3rd ed.: Oxford University Press.","edition":"3"},{"key":"e_1_3_4_31_1","doi-asserted-by":"crossref","unstructured":"Jiang Bowen Xie Yangxinyu Wang Xiaomeng Yuan Yuan Hao Zhuoqun Bai Xinyi Su Weijie J. Taylor Camillo Jose Mallick Tanwi. 2025. \u201cTowards Rationality in Language and Multimodal Agents: A Survey.\u201d Pp. 3656\u20133675 in Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers). Association for Computational Linguistics. doi:10.18653\/v1\/2025.naacl-long.186.","DOI":"10.18653\/v1\/2025.naacl-long.186"},{"key":"e_1_3_4_32_1","doi-asserted-by":"crossref","unstructured":"Julien Boelaert Coavoux Samuel Ollion \u00c9tienne Petev Ivaylo Pr\u00e4g Patrick. 2025b. \u201cReplication Materials to: Machine Bias. How Do Generative Language Models Answer Opinion Polls?.\u201d doi:10.17605\/OSF.IO\/YZECT.","DOI":"10.31235\/osf.io\/r2pnb_v2"},{"key":"e_1_3_4_33_1","doi-asserted-by":"publisher","DOI":"10.1177\/00491241251337316"},{"key":"e_1_3_4_34_1","doi-asserted-by":"crossref","unstructured":"Liu Jiaxin Tang Yixuan Yang Yi Tam Kar Yan. 2025. \u201cEvaluating and Aligning Human Economic Risk Preferences in LLMs.\u201d Pp. 18174\u201318188 in Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing. Association for Computational Linguistics. doi:10.18653\/v1\/2025.emnlp-main.917.","DOI":"10.18653\/v1\/2025.emnlp-main.917"},{"key":"e_1_3_4_35_1","doi-asserted-by":"publisher","DOI":"10.1162\/tacl_a_00638"},{"key":"e_1_3_4_36_1","unstructured":"Liu Yinhong Guo Zhijiang Liang Tianya Shareghi Ehsan Vuli\u0107 Ivan Collier Nigel. 2024. \u201cAligning With Logic: Measuring Evaluating and Improving Logical Preference Consistency in Large Language Models.\u201d arXiv preprint arXiv:2410.02205."},{"key":"e_1_3_4_37_1","doi-asserted-by":"publisher","DOI":"10.1177\/00491241251342008"},{"key":"e_1_3_4_38_1","doi-asserted-by":"publisher","DOI":"10.52202\/068431-2011"},{"key":"e_1_3_4_39_1","unstructured":"Packer Charles Wooders Sarah Lin Kevin Fang Vivian Patil Shishir G. Stoica Ion Gonzalez Joseph E.. 2023. \u201cMemGPT: Towards LLMs as Operating Systems.\u201d arXiv preprint arXiv:2310.08560."},{"key":"e_1_3_4_40_1","doi-asserted-by":"crossref","unstructured":"Park Sung Joon O\u2019Brien Joseph C. Cai Carrie J. Morris Meredith Ringel Liang Percy Bernstein Michael S.. 2023. \u201cGenerative Agents: Interactive Simulacra of Human Behavior.\u201d Pp. 1\u201322 in Proceedings of the 36th Annual ACM Symposium on User Interface Software and Technology. ACM. doi:10.1145\/3586183.3606763.","DOI":"10.1145\/3586183.3606763"},{"key":"e_1_3_4_41_1","unstructured":"Raman Narun Krishnamurthi Lundy Taylor Amouyal Samuel Joseph Levine Yoav Leyton-Brown Kevin Tennenholtz Moshe. 2024. \u201cSTEER: Assessing the Economic Rationality of Large Language Models.\u201d Pp. 42026\u201342047 in Proceedings of the 41st International Conference on Machine Learning Vol. 235 of Proceedings of Machine Learning Research. PMLR."},{"key":"e_1_3_4_42_1","doi-asserted-by":"crossref","unstructured":"Ross Jillian Kim Yoon Lo Andrew W.. 2024. \u201cLLM Economicus? Mapping the Behavioral Biases of LLMs via Utility Theory.\u201d in Proceedings of the First Conference on Language Modeling (COLM). https:\/\/openreview.net\/forum?id=Rx3wC8sCTJ.","DOI":"10.2139\/ssrn.4926791"},{"key":"e_1_3_4_43_1","doi-asserted-by":"publisher","DOI":"10.2307\/2548836"},{"key":"e_1_3_4_44_1","doi-asserted-by":"publisher","DOI":"10.1037\/0003-066X.54.2.93"},{"key":"e_1_3_4_45_1","unstructured":"Sclar Melanie Choi Yejin Tsvetkov Yulia Suhr Alane. 2024. \u201cQuantifying Language Models\u2019 Sensitivity to Spurious Features in Prompt Design Or: How I Learned to Start Worrying About Prompt Formatting.\u201d in Proceedings of the 12th International Conference on Learning Representations."},{"key":"e_1_3_4_46_1","unstructured":"T\u00f6rnberg Petter Valeeva Diliara Uitermark Justus Bail Christopher. 2023. \u201cSimulating Social Media Using Large Language Models to Evaluate Alternative News Feed Algorithms.\u201d doi:10.48550\/arXiv.2310.05984."},{"key":"e_1_3_4_47_1","doi-asserted-by":"publisher","DOI":"10.2307\/1912771"},{"key":"e_1_3_4_48_1","doi-asserted-by":"publisher","DOI":"10.1016\/0304-4076(90)90051-T"},{"key":"e_1_3_4_49_1","unstructured":"Wang Xuezhi Wei Jason Schuurmans Dale Le Quoc Chi Ed Narang Sharan Chowdhery Aakanksha Zhou Denny. 2023. \u201cSelf-Consistency Improves Chain of Thought Reasoning in Language Models.\u201d in Proceedings of the 11th International Conference on Learning Representations."},{"key":"e_1_3_4_50_1","doi-asserted-by":"crossref","unstructured":"Zhao Xiutian Wang Ke Peng Wei. 2024. \u201cMeasuring the Inconsistency of Large Language Models in Preferential Ranking.\u201d in Proceedings of the 1st Workshop on Towards Knowledgeable Language Models (KnowLLM). Association for Computational Linguistics. https:\/\/arxiv.org\/abs\/2410.08851.","DOI":"10.18653\/v1\/2024.knowllm-1.14"},{"key":"e_1_3_4_51_1","doi-asserted-by":"publisher","DOI":"10.1162\/coli_a_00502"}],"container-title":["Sociological Methods &amp; Research"],"language":"en","link":[{"URL":"https:\/\/journals.sagepub.com\/doi\/pdf\/10.1177\/00491241261492203","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/journals.sagepub.com\/doi\/full-xml\/10.1177\/00491241261492203","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/journals.sagepub.com\/doi\/pdf\/10.1177\/00491241261492203","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,2]],"date-time":"2026-10-02T13:50:36Z","timestamp":1790949036000},"score":58.995594,"resource":{"primary":{"URL":"https:\/\/journals.sagepub.com\/doi\/10.1177\/00491241261492203"}},"issued":{"date-parts":[[2026,9,29]]},"references-count":50,"alternative-id":["10.1177\/00491241261492203"],"URL":"https:\/\/doi.org\/10.1177\/00491241261492203","ISSN":["0049-1241","1552-8294"],"issn-type":[{"value":"0049-1241","type":"print"},{"value":"1552-8294","type":"electronic"}],"published":{"date-parts":[[2026,9,29]]},"article-number":"00491241261492203"},{"indexed":{"date-parts":[[2026,9,12]],"date-time":"2026-09-12T10:07:26Z","timestamp":1789207646463,"version":"build-2803163510"},"posted":{"date-parts":[[2026,9,11]]},"group-title":"EdArXiv","reference-count":0,"publisher":"Center for Open Science","license":[{"start":{"date-parts":[[2026,9,11]],"date-time":"2026-09-11T00:00:00Z","timestamp":1789084800000},"content-version":"unspecified","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by-nc-nd\/4.0\/legalcode"}],"content-domain":{"domain":[],"crossmark-restriction":false},"abstract":"<p>Large language models (LLMs) can now supply complete solutions to most academic tasks on demand. Whether that helps or harms learning depends on what the interaction does to the effortful processing that learning requires. This article proposes the Neuroplasticity-AI Interaction Model (NAIM), which distinguishes three interaction pathways: *direct bypass*, in which the model supplies the solution and the learner's generative effort is eliminated; cognitive offloading, in which specific subprocesses are delegated and the effect depends on whether those subprocesses are the learning target; and scaffolded struggle, in which the model supplies hints, questions, and feedback while the learner retains the generative work. Drawing on complementary learning systems theory, cognitive load theory, and research on desirable difficulties and productive failure, the model explains why recent randomized studies find both harm and benefit from the same underlying technology: harm follows bypass, benefit follows scaffolding. The model yields eight testable propositions concerning transfer, retention, the dissociation between learner preference and learning, moderating effects of prior knowledge and metacognition, and dose-dependent accumulation of deficits. A research agenda specifies the designs and outcome measures needed to test each proposition. Implications for tutor design, assessment, and policy follow from the propositions rather than preceding them.<\/p>","DOI":"10.35542\/osf.io\/8f376_v1","type":"posted-content","created":{"date-parts":[[2026,9,12]],"date-time":"2026-09-12T09:30:22Z","timestamp":1789205422000},"source":"Crossref","is-referenced-by-count":0,"title":["Bypass, Offload, or Scaffold: A Conceptual Model of How Large Language Models Shape Learning"],"prefix":"10.35542","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-1209-8565","authenticated-orcid":true,"given":"Ernesto","family":"Lee","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"15934","deposited":{"date-parts":[[2026,9,12]],"date-time":"2026-09-12T09:30:22Z","timestamp":1789205422000},"score":29.052929,"resource":{"primary":{"URL":"https:\/\/osf.io\/8f376_v1"}},"issued":{"date-parts":[[2026,9,11]]},"references-count":0,"URL":"https:\/\/doi.org\/10.35542\/osf.io\/8f376_v1","published":{"date-parts":[[2026,9,11]]},"subtype":"preprint"},{"indexed":{"date-parts":[[2025,10,24]],"date-time":"2025-10-24T16:49:57Z","timestamp":1761324597572,"version":"3.28.0"},"reference-count":37,"publisher":"IEEE","license":[{"start":{"date-parts":[[2024,6,30]],"date-time":"2024-06-30T00:00:00Z","timestamp":1719705600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2024,6,30]],"date-time":"2024-06-30T00:00:00Z","timestamp":1719705600000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"content-domain":{"domain":[],"crossmark-restriction":false},"published-print":{"date-parts":[[2024,6,30]]},"DOI":"10.1109\/ijcnn60899.2024.10650668","type":"proceedings-article","created":{"date-parts":[[2024,9,9]],"date-time":"2024-09-09T17:35:05Z","timestamp":1725903305000},"page":"1-9","source":"Crossref","is-referenced-by-count":1,"title":["Exploring and Improving Consistency in Large Language Models for Multiple-Choice Question Assessment"],"prefix":"10.1109","author":[{"given":"Wenjie","family":"Zhou","sequence":"first","affiliation":[{"name":"Soochow University,School of Computer Science and Technology,Soochow,China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Xiangyu","family":"Duan","sequence":"additional","affiliation":[{"name":"Soochow University,School of Computer Science and Technology,Soochow,China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"263","reference":[{"key":"ref1","first-page":"1877","article-title":"Language models are few-shot learners[J]","volume":"33","author":"Brown","year":"2020","journal-title":"Advances in neural information processing systems"},{"journal-title":"Llama 2: Open foundation and fine-tuned chat models[J]","year":"2023","author":"Touvron","key":"ref2"},{"key":"ref3","first-page":"27730","article-title":"Training language models to follow instructions with human feedback[J]","volume":"35","author":"Ouyang","year":"2022","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref4","first-page":"3","volume-title":"ICML 2011 Workshop on New Developments in Imitation Learning (July 2011)","volume":"855","author":"Knox"},{"journal-title":"Proximal policy optimization algorithms[J]","year":"2017","author":"Schulman","key":"ref5"},{"key":"ref6","first-page":"301","article-title":"Why can GPT learn in-context? language models secretly perform gradient descent as metaoptimizers[C]","volume-title":"Findings of the Association for Computational Linguistics: ACL 2023. 2023: 4005-4019. August 1987 [Digests 9th Annual Conf. Magnetics Japan","author":"Dai"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1145\/3560815"},{"journal-title":"Glm-130b: An open bilingual pretrained model[J]","year":"2022","author":"Zeng","key":"ref8"},{"journal-title":"Measuring massive multitask language understanding[J]","year":"2020","author":"Hendrycks","key":"ref9"},{"journal-title":"C-eval: A multi-level multidiscipline chinese evaluation suite for foundation models[J]","year":"2023","author":"Huang","key":"ref10"},{"key":"ref11","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2022.acl-long.234"},{"key":"ref12","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.findings-naacl.149"},{"volume-title":"Training verifiers to solve math word problems, 2021[J]","year":"2021","author":"Cobbe","key":"ref13"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.findings-acl.824"},{"key":"ref15","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2022.acl-long.229"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2023.emnlp-main.991"},{"journal-title":"Evaluating task understanding through multilingual consistency: A ChatGPT case study[J]","year":"2023","author":"Ohmer","key":"ref17"},{"journal-title":"Judging LLM-as-ajudge with MT-Bench and Chatbot Arena[J]","year":"2023","author":"Zheng","key":"ref18"},{"key":"ref19","doi-asserted-by":"publisher","DOI":"10.1145\/3641289"},{"key":"ref20","doi-asserted-by":"publisher","DOI":"10.1162\/tacl_x_00455"},{"journal-title":"Assessing Hidden Risks of LLMs: An Empirical Study on Robustness, Consistency, and Credibility[J]","year":"2023","author":"Ye","key":"ref21"},{"journal-title":"Evaluating the Decency and Consistency of Data Validation Tests Generated by LLMs[J]","year":"2023","author":"Alexander","key":"ref22"},{"journal-title":"A prompt pattern catalog to enhance prompt engineering with chatgpt[J]","year":"2023","author":"White","key":"ref23"},{"article-title":"Large language models are not robust multiple choice selectors[C]","volume-title":"The Twelfth International Conference on Learning Representations","author":"Zheng","key":"ref24"},{"key":"ref25","first-page":"24824","article-title":"Chain-of-thought prompting elicits reasoning in large language models[J]","volume":"35","author":"Wei","year":"2022","journal-title":"Advances in neural information processing systems"},{"key":"ref26","first-page":"22199","article-title":"Large language models are zero-shot reasoners[J]","volume":"35","author":"Kojima","year":"2022","journal-title":"Advances in neural information processing systems"},{"journal-title":"\u201cMy Answer is C\u201d: First-Token Probabilities Do Not Match Text Answers in Instruction-Tuned Language Models[J]","year":"2024","author":"Wang","key":"ref27"},{"journal-title":"Self-instruct: Aligning language models with self-generated instructions[J]","year":"2022","author":"Wang","key":"ref28"},{"journal-title":"Dola: Decoding by contrasting layers improves factuality in large language models[J]","year":"2023","author":"Chuang","key":"ref29"},{"journal-title":"Opt-iml: Scaling language model instruction meta learning through the lens of generalization[J]","year":"2022","author":"Iyer","key":"ref30"},{"key":"ref31","first-page":"36","article-title":"Principle-driven self-alignment of language models from scratch with minimal human supervision[J]","author":"Sun","year":"2024","journal-title":"Advances in Neural Information Processing Systems"},{"key":"ref32","doi-asserted-by":"publisher","DOI":"10.1162\/tacl_x_00455"},{"journal-title":"Platypus: Quick, cheap, and powerful refinement of llms[J]","year":"2023","author":"Lee","key":"ref33"},{"journal-title":"Lora: Low-rank adaptation of large language models[J]","year":"2021","author":"Hu","key":"ref34"},{"journal-title":"Qwen-vl: A frontier large vision-language model with versatile abilities[J]","year":"2023","author":"Bai","key":"ref35"},{"journal-title":"Baichuan 2: Open large-scale language models[J]","year":"2023","author":"Yang","key":"ref36"},{"journal-title":"Holistic evaluation of language models[J]","year":"2022","author":"Liang","key":"ref37"}],"event":{"name":"2024 International Joint Conference on Neural Networks (IJCNN)","start":{"date-parts":[[2024,6,30]]},"location":"Yokohama, Japan","end":{"date-parts":[[2024,7,5]]}},"container-title":["2024 International Joint Conference on Neural Networks (IJCNN)"],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/10649807\/10649898\/10650668.pdf?arnumber=10650668","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,9,10]],"date-time":"2024-09-10T05:42:41Z","timestamp":1725946961000},"score":29.04638,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/10650668\/"}},"issued":{"date-parts":[[2024,6,30]]},"references-count":37,"URL":"https:\/\/doi.org\/10.1109\/ijcnn60899.2024.10650668","published":{"date-parts":[[2024,6,30]]}},{"indexed":{"date-parts":[[2026,5,30]],"date-time":"2026-05-30T14:01:42Z","timestamp":1780149702081,"version":"3.54.0"},"posted":{"date-parts":[[2026]]},"group-title":"SSRN","reference-count":0,"publisher":"Elsevier BV","content-domain":{"domain":[],"crossmark-restriction":false},"abstract":"<jats:p>Large language models (LLMs) are emerging as new touchpoints in the customer journey, influencing how consumers discover and learn about brands. The authors argue that in these environments, brand knowledge is no longer formed solely in consumer memory but is also constructed and conveyed by LLMs. As firms continuously seek to shift brand-related associations, this research examines how brand knowledge updates in LLMs. Compared to consumer-based updating, LLMs differ structurally at the information-exposure stage: whereas consumers encounter brand information through diverse and partly stochastic touchpoints, LLMs rely on structured retrieval initiated by user queries. Consequently, the informational basis for brand-knowledge updating becomes contingent on query formulation and retrieval design, leading to systematic variation in how brand knowledge is conveyed across and within LLMs. The authors test these predictions in a controlled synthetic field experiment that leverages strategic brand repositioning as an exogenous shift in brand associations for recently established brands and tracks conveyed knowledge in model responses over time. The results provide broad support for retrieval-driven updating, with systematic heterogeneity across models and query types. The authors discuss implications for brand knowledge theory and identify retrievability as a key managerial lever in AI-mediated customer journeys.<\/jats:p>","DOI":"10.2139\/ssrn.6798479","type":"posted-content","created":{"date-parts":[[2026,5,30]],"date-time":"2026-05-30T13:18:24Z","timestamp":1780147104000},"source":"Crossref","is-referenced-by-count":0,"title":["Beyond Consumer Memory: How Large Language Models Update Brand Knowledge"],"prefix":"10.2139","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-7423-8589","authenticated-orcid":true,"given":"Andreas","family":"Hamann","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Florian","family":"Stahl","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"P. K.","family":"Kannan","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"78","deposited":{"date-parts":[[2026,5,30]],"date-time":"2026-05-30T13:18:24Z","timestamp":1780147104000},"score":28.58923,"resource":{"primary":{"URL":"https:\/\/www.ssrn.com\/abstract=6798479"}},"issued":{"date-parts":[[2026]]},"references-count":0,"URL":"https:\/\/doi.org\/10.2139\/ssrn.6798479","published":{"date-parts":[[2026]]},"subtype":"preprint"},{"indexed":{"date-parts":[[2026,10,26]],"date-time":"2026-10-26T13:13:22Z","timestamp":1793020402660,"version":"4.5.0"},"reference-count":0,"publisher":"Association for the Advancement of Artificial Intelligence (AAAI)","issue":"18","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["AAAI"],"abstract":"<jats:p>Accurately gauging the confidence level of Large Language Models' (LLMs) predictions is pivotal for their reliable application. However, LLMs are often uncalibrated inherently and elude conventional calibration techniques due to their proprietary nature and massive scale. In this work, we derive model confidence from the distribution of multiple randomly sampled generations, using three measures of consistency. We extensively evaluate eleven open and closed-source models on nine reasoning datasets. Results show that consistency-based calibration methods outperform existing post-hoc approaches in terms of calibration error. Meanwhile, we find that factors such as intermediate explanations, model scaling, and larger sample sizes enhance calibration, while instruction-tuning makes calibration more difficult. Moreover, confidence scores obtained from consistency can potentially enhance model performance. Finally, we offer guidance on choosing suitable consistency metrics for calibration, tailored to model characteristics such as the exposure to instruction-tuning and RLHF.<\/jats:p>","DOI":"10.1609\/aaai.v39i18.34120","type":"journal-article","created":{"date-parts":[[2025,4,11]],"date-time":"2025-04-11T08:52:46Z","timestamp":1744361566000},"page":"19260-19268","source":"Crossref","is-referenced-by-count":19,"title":["Calibrating Large Language Models with Sample Consistency"],"prefix":"10.1609","volume":"39","author":[{"given":"Qing","family":"Lyu","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kumar","family":"Shridhar","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Chaitanya","family":"Malaviya","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Li","family":"Zhang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yanai","family":"Elazar","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Niket","family":"Tandon","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Marianna","family":"Apidianaki","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mrinmaya","family":"Sachan","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Chris","family":"Callison-Burch","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"9382","published-online":{"date-parts":[[2025,4,11]]},"container-title":["Proceedings of the AAAI Conference on Artificial Intelligence"],"link":[{"URL":"https:\/\/ojs.aaai.org\/index.php\/AAAI\/article\/download\/34120\/36275","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/ojs.aaai.org\/index.php\/AAAI\/article\/download\/34120\/36275","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,4,11]],"date-time":"2025-04-11T08:52:46Z","timestamp":1744361566000},"score":28.557348,"resource":{"primary":{"URL":"https:\/\/ojs.aaai.org\/index.php\/AAAI\/article\/view\/34120"}},"issued":{"date-parts":[[2025,4,11]]},"references-count":0,"journal-issue":{"issue":"18","published-online":{"date-parts":[[2025,4,11]]}},"URL":"https:\/\/doi.org\/10.1609\/aaai.v39i18.34120","ISSN":["2374-3468","2159-5399"],"issn-type":[{"value":"2374-3468","type":"electronic"},{"value":"2159-5399","type":"print"}],"published":{"date-parts":[[2025,4,11]]}}],"items-per-page":5,"query":{"start-index":0,"search-terms":null}}}