{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,25]],"date-time":"2026-03-25T15:18:21Z","timestamp":1774451901220,"version":"3.50.1"},"reference-count":61,"publisher":"Springer Science and Business Media LLC","issue":"2","license":[{"start":{"date-parts":[[2024,5,31]],"date-time":"2024-05-31T00:00:00Z","timestamp":1717113600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.springernature.com\/gp\/researchers\/text-and-data-mining"},{"start":{"date-parts":[[2024,5,31]],"date-time":"2024-05-31T00:00:00Z","timestamp":1717113600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/www.springernature.com\/gp\/researchers\/text-and-data-mining"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Ethics Inf Technol"],"published-print":{"date-parts":[[2024,6]]},"DOI":"10.1007\/s10676-024-09779-1","type":"journal-article","created":{"date-parts":[[2024,5,31]],"date-time":"2024-05-31T02:01:30Z","timestamp":1717120890000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":5,"title":["Getting it right: the limits of fine-tuning large language models"],"prefix":"10.1007","volume":"26","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-3102-5932","authenticated-orcid":false,"given":"Jacob","family":"Browning","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"297","published-online":{"date-parts":[[2024,5,31]]},"reference":[{"key":"9779_CR1","unstructured":"Agarwal, C., Tanneru, S. H., & Lakkaraju, H. (2024). Faithfulness vs. Plausibility: On the (Un) reliability of explanations from large language models. arXiv Preprint arXiv, 2402, 04614."},{"key":"9779_CR3","doi-asserted-by":"crossref","unstructured":"Andreas, J. (2022). Language models as agent models. arXiv preprint:arXiv, 2212, 01681.","DOI":"10.18653\/v1\/2022.findings-emnlp.423"},{"key":"9779_CR4","doi-asserted-by":"publisher","DOI":"10.48550\/arxiv.2212.08073","author":"Y Bai","year":"2022","unstructured":"Bai, Y., Kadavath, S., Kundu, S., Askell, A., Kernion, J., Jones, A., et al. (2022). Constitutional AI: Harmlessness from AI feedback. Arxiv. https:\/\/doi.org\/10.48550\/arxiv.2212.08073.","journal-title":"Arxiv"},{"key":"9779_CR5","unstructured":"Belanger, A. (7\/7\/2023). ChatGPT usage drop for the first time as users turn to uncensored chatbots. Ars Technica."},{"key":"9779_CR6","doi-asserted-by":"publisher","unstructured":"Bender, E. M., Gebru, T., Mcmillan-Major, A., & Shmitchell, S. (2021). On the dangers of stochastic parrots: Can language models be too big? Proceedings of the 2021 ACM Conference on Fairness, Accountability, and Transparency, https:\/\/doi.org\/10.1145\/3442188.3445922.","DOI":"10.1145\/3442188.3445922"},{"key":"9779_CR8","unstructured":"Brown, T., Mann, B., Ryder, N., Subbiah, M., Kaplan, J. D., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A., Agarwal, S., Herbert-Voss, A., Krueger, G., Henighan, T., Child, R., Ramesh, A., Ziegler, D., Wu, J., Winter, C., & Amodei, D. (2020). Language models are few-shot learners. Advances in Neural Information Processing Systems."},{"key":"9779_CR9","unstructured":"Browning, J., & LeCun, Y. (2022, August 22). AI and the Limits of Language. NOEMA. https:\/\/www.noemamag.com\/ai-and-the-limits-of-language\/."},{"key":"9779_CR10","doi-asserted-by":"publisher","DOI":"10.1007\/s00146-023-01724-y","author":"J Browning","year":"2023","unstructured":"Browning, J., (2023). Personhood why large language models don\u2019t understand us. AI & Soc. https:\/\/doi.org\/10.1007\/s00146-023-01724-y.","journal-title":"AI & Soc"},{"key":"9779_CR11","unstructured":"Bump, P. (2024, January 17). Doing your own research is a good way to end up being wrong \u2013 The Washington Post. Washington Post, . www.washingtonpost.com\/politics\/2024\/01\/17\/do-your-own-research-study\/."},{"key":"9779_CR12","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2307.15217","author":"S Casper","year":"2023","unstructured":"Casper, S., Davies, X., Shi, C., Gilbert, T., Scheurer, J., Rando, J., Freedman, R., Korbak, T., Lindner, D., Freire, P., Wang, T., Marks, S., Segerie, C., Carroll, M., Peng, A., Christoffersen, P., Damani, M., Slocum, S., Anwar, U., & Hadfield-Menell, D. (2023). Open problems and fundamental limitations of reinforcement learning from human feedback. Arxiv. https:\/\/doi.org\/10.48550\/arXiv.2307.15217.","journal-title":"Arxiv"},{"key":"9779_CR13","unstructured":"Chen, L., Zaharia, M., & Zou, J. (2023). How is ChatGPT\u2019s behavior changing over time? arXiv Preprint arXiv, 2307, 09009."},{"key":"9779_CR130","doi-asserted-by":"publisher","unstructured":"Contreras Kallens, P., Kristensen-McLachlan, R. D., Christiansen M. H. (2023). Large language models demonstrate the potential of statistical learning in language. Cognitive Science, 47(3), e13256. https:\/\/doi.org\/10.1111\/cogs.13256. PMID: 36840975.","DOI":"10.1111\/cogs.13256"},{"key":"9779_CR14","unstructured":"Coelho Mollo, D. (2023). R. Milli\u00e8re (Ed.), The vector grounding problem. arXiv Preprint: arXiv, 2304, 01481."},{"key":"9779_CR16","doi-asserted-by":"crossref","unstructured":"Deng, G., Liu, Y., Li, Y., Wang, K., Zhang, Y., Li, Z., Wang, H., Zhang, T., & Liu, Y. (2023). Jailbreaker: Automated jailbreak across multiple large language model chatbots. arXiv Preprint arXiv, 2307, 08715.","DOI":"10.14722\/ndss.2024.24188"},{"key":"9779_CR17","doi-asserted-by":"publisher","DOI":"10.48550\/arxiv.2302.12313","author":"V Dentella","year":"2023","unstructured":"Dentella, V., Murphy, E., Marcus, G., & Leivada, E. (2023). Testing AI performance on less frequent aspects of language reveals insensitivity to underlying meaning. Arxiv. https:\/\/doi.org\/10.48550\/arxiv.2302.12313.","journal-title":"Arxiv"},{"issue":"10","key":"9779_CR18","doi-asserted-by":"publisher","first-page":"1066","DOI":"10.1162\/tacl_a_00506","volume":"2022","author":"N Dziri","year":"2022","unstructured":"Dziri, N., Rashkin, H., Linzen, T., & Reitter, D. (2022). Evaluating attribution in dialogue systems: The BEGIN Benchmark. Transactions of the Association for Computational Linguistics, 2022(10), 1066\u20131083. https:\/\/doi.org\/10.1162\/tacl_a_00506.","journal-title":"Transactions of the Association for Computational Linguistics"},{"key":"9779_CR19","unstructured":"El-Mahdi, E., Farhadkhani, S., Guerraoui, R., Gupta, N. (2022). L\u00ea-Nguy\u00ean Hoang, Pinot, R.,. SoK: On the impossible security of very large foundation models. Arxiv."},{"key":"9779_CR21","doi-asserted-by":"publisher","DOI":"10.48550\/arxiv.2302.07459","author":"D Ganguli","year":"2023","unstructured":"Ganguli, D., Askell, A., Schiefer, N., Liao, T. I., Luko\u0161i\u016bt\u0117, K., Chen, A., et al. (2023). The capacity for moral self-correction in large language models. Arxiv. https:\/\/doi.org\/10.48550\/arxiv.2302.07459.","journal-title":"Arxiv"},{"key":"9779_CR22","unstructured":"Ghaffery, S. (2023, September 20). ChatGPT usage is rising again as students return to school. Bloomberg. https:\/\/www.bloomberg.com\/news\/articles\/2023-09-20\/chatgpt-usage-is-rising-again-as-students-return-to-school#xj4y7vzkg."},{"key":"9779_CR23","doi-asserted-by":"crossref","unstructured":"Grice, P. (1975). Logic and conversation. In Cole, P.; Morgan, J. (Eds.). Syntax and semantics. Vol. 3: Speech acts. New York: Academic Press.","DOI":"10.1163\/9789004368811_003"},{"key":"9779_CR27","unstructured":"Hoel, E. (2024, March 29). A.I.-Generated garbage is polluting our culture. The New York Times, The New York Times. www.nytimes.com\/2024\/03\/29\/opinion\/ai-internet-x-youtube.html."},{"key":"9779_CR26","unstructured":"Hofstadter, D., & G\u00f6del Escher, Bach, and Ai (2023, July 10). The Atlantic, Atlantic media company, www.theatlantic.com\/ideas\/archive\/2023\/07\/godel-escher-bach-geb-ai\/674589\/."},{"key":"9779_CR29","unstructured":"Jain, N., Han, K., Gu, A., Li, W. D., Yan, F., Zhang, T., & Stoica, I. (2024). LiveCodeBench: Holistic and contamination free evaluation of large language models for code. arXiv Preprint: arXiv, 2403, 07974."},{"key":"9779_CR30","doi-asserted-by":"publisher","unstructured":"Kalai, A., & Vempala, S. (2023). Calibrated language models must hallucinate. Arxiv. https:\/\/doi.org\/10.48550\/arXiv.2311.14648.","DOI":"10.48550\/arXiv.2311.14648"},{"key":"9779_CR31","doi-asserted-by":"crossref","unstructured":"Kasirzadeh, A., & Gabriel, I. (2023). In conversation with artificial intelligence: Aligning language models with human values. Philosophy of Technology.","DOI":"10.1007\/s13347-023-00606-x"},{"key":"9779_CR32","unstructured":"Kempt, H., Lavie, A., & Lavie, S. K. (2023). Appropriateness is all you need! ArXiv, 1\u201325."},{"key":"9779_CR33","doi-asserted-by":"publisher","DOI":"10.2139\/ssrn.4372889","author":"J Koco\u0144","year":"2023","unstructured":"Koco\u0144, J., Cichecki, I., Kaszyca, O., Kochanek, M., Szyd\u0142o, D., Baran, J., et al. (2023). Chatgpt: Jack of all trades, master of none. Elsevier BV. https:\/\/doi.org\/10.2139\/ssrn.4372889.","journal-title":"Elsevier BV"},{"key":"9779_CR34","doi-asserted-by":"publisher","unstructured":"Korbak, T., Shi, K., Chen, A., Bhalerao, R., Buckley, C. L., Phang, J. (2023). Pretraining language models with human preferences. Arxiv. https:\/\/doi.org\/10.48550\/arxiv.2302.08582.","DOI":"10.48550\/arxiv.2302.08582"},{"key":"9779_CR35","unstructured":"Kumar, A., Raghunathan, A., Jones, R., Ma, T., & Liang, P. (2022). Fine-tuning can distort pretrained features and underperform out-of-distribution. arXiv Preprint arXiv, 220210054."},{"key":"9779_CR36","unstructured":"LeCun, Y. (2022). A path towards autonomous machine intelligence, Version 0.9.2, 2022-06-27. https:\/\/openreview.net\/forum?id=BZ5a1r-kVsf."},{"key":"9779_CR37","doi-asserted-by":"publisher","first-page":"157","DOI":"10.1162\/tacl_a_00638","volume":"12","author":"NF Liu","year":"2024","unstructured":"Liu, N. F., Lin, K., Hewitt, J., Paranjape, A., Bevilacqua, M., Petroni, F., & Liang, P. (2024). Lost in the middle: How language models use long contexts. Transactions of the Association for Computational Linguistics, 12, 157\u2013173.","journal-title":"Transactions of the Association for Computational Linguistics"},{"key":"9779_CR39","unstructured":"Mahowald, K., Ivanova, A., Fedorenko, E., Blank, I. A., Tenenbaum, J., & Kanwisher, N. (2024, January 24). Google\u2019s powerful AI spotlights a human cognitive glitch: Mistaking fluent speech for fluent thought. The Conversation. https:\/\/theconversation.com\/googles-powerful-ai-spotlights-a-human-cognitive-glitch-mistaking-fluent-speech-for-fluent-thought-185099."},{"key":"9779_CR38","unstructured":"McCoy, R. T., Yao, S., Friedman, D., Hardy, M., & Griffiths, T. L. (2023). Embers of autoregression: Understanding large language models through the problem they are trained to solve. arXiv Preprint arXiv:230913638."},{"key":"9779_CR40","unstructured":"Milliere, R. (2020, September 09). Welcome to the next level of bullshit. Nautilus. https:\/\/nautil.us\/welcome-to-the-next-level-of-bullshit-9245\/."},{"key":"9779_CR41","unstructured":"Moderator (2022). Temporary Policy: Generative AI (e.g., Chatgpt) is banned. Meta Stack Overflow, 5 Dec. meta.stackoverflow.com\/questions\/421831\/temporary-policy-generative-ai-e-g-chatgpt-is-banned."},{"key":"9779_CR42","unstructured":"Offenhartz, J. (2024). NYC\u2019s AI chatbot was caught telling businesses to break the law. The City Isn\u2019t Taking It Down. AP News, AP News, 4 Apr. apnews.com\/article\/new-york-city-chatbot-misinformation-6ebc71db5b770b9969c906a7ee4fae21."},{"key":"9779_CR43","unstructured":"Orf, D. (2023, March 1). Elon musk building anti-woke ai to rival ChatGPT: What it means. Retrieved from Popular Mechanics: https:\/\/www.popularmechanics.com\/technology\/robots\/a43126181\/elon-musk-anti-woke-artificial-intelligence\/."},{"key":"9779_CR44","unstructured":"Ortega, P. A., Kunesch, M., Del\u00e9tang, G., Genewein, T., Grau-Moya, J., Veness, J., & Legg, S. (2021). Shaking the foundations: Delusions in sequence models for interaction and control. arXiv Preprint arXiv, 2110., 10819."},{"issue":"1","key":"9779_CR45","doi-asserted-by":"publisher","first-page":"447","DOI":"10.1146\/annurev-linguistics-031120-122924","volume":"8","author":"E Pavlick","year":"2022","unstructured":"Pavlick, E. (2022). Semantic structure in deep learning. Annu Rev Linguist, 8(1), 447\u2013471.","journal-title":"Annu Rev Linguist"},{"key":"9779_CR46","unstructured":"Piantasodi, S. T., & Hill, F. (2022). Meaning without reference in large language models, ArXiv, pp.1\u20138."},{"key":"9779_CR47","unstructured":"Piltch, A. (2023, 06\/11). Plagiarism Engine: Google\u2019s Content-Swiping AI Could Break the Internet. https:\/\/www.tomshardware.com\/news\/google-sge-break-internet."},{"key":"9779_CR48","doi-asserted-by":"crossref","unstructured":"Razeghi, Y., Logan, I. V., Gardner, R. L., M., & Singh, S. (2022). Impact of pretraining term frequencies on few-shot reasoning. arXiv Preprint: arXiv, 220207206.","DOI":"10.18653\/v1\/2022.findings-emnlp.59"},{"key":"9779_CR49","unstructured":"Russell, S. (2019). Human-Compatible AI, Viking."},{"key":"9779_CR50","unstructured":"Schank, R., & Abelson, R. P. (1977). Scripts, plans, goals and understanding: An inquiry into human knowledge structures. New Jersey: Erlbaum. ISBN 0-470-99033-3."},{"key":"9779_CR51","unstructured":"Scott-Philips, T. (2014). Speaking our minds: Why human communication is different, and how language evolved to make it special. Red Globe."},{"issue":"2","key":"9779_CR52","doi-asserted-by":"publisher","first-page":"341","DOI":"10.1007\/s11023-022-09602-0","volume":"32","author":"A Sobieszek","year":"2022","unstructured":"Sobieszek, A., & Price, T. (2022). Playing games with Ais: The limits of GPT-3 and similar large Language models. Minds and Machines, 32(2), 341\u2013364.","journal-title":"Minds and Machines"},{"key":"9779_CR53","doi-asserted-by":"crossref","unstructured":"Sutton, R., & Barto, A. (1998). Reinfocement Learning: An introduction. MIT Press.","DOI":"10.1109\/TNN.1998.712192"},{"key":"9779_CR54","unstructured":"Tangerman, V. (2023, February 15). Microsoft\u2019s Bing AI Now Threatening Users Who Provoke It. Retrieved from Futurism: https:\/\/futurism.com\/microsoft-bing-ai-threatening."},{"key":"9779_CR55","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2201.08239","author":"R Thoppilan","year":"2022","unstructured":"Thoppilan, R., De Freitas, D., Hall, J., Shazeer, N., Kulshreshtha, A., Cheng, H., et al. (2022). Lamda: Language models for dialog applications. Arxiv. https:\/\/doi.org\/10.48550\/arXiv.2201.08239.","journal-title":"Arxiv"},{"key":"9779_CR57","unstructured":"Valmeekam, K., Olmo, A., Sreeharan, S., & Kambhampati, S. (2023). Large Language models still can\u2019t plan (pp. 1\u201321). Arxiv."},{"key":"9779_CR56","unstructured":"Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017). Attention is all you need. Advances in Neural Information Processing Systems 30, Long Beach, CA."},{"key":"9779_CR58","unstructured":"Wallace, R. J. (1994). Responsibility and the Moral sentiments. Harvard University Press."},{"key":"9779_CR59","doi-asserted-by":"publisher","unstructured":"Welbl, J., Glaese, A., Uesato, J., Dathathri, S., Mellor, J., Hendricks, L. A. (2021). Challenges in detoxifying language models Association for Computational Linguistics. https:\/\/doi.org\/10.18653\/v1\/2021.findings-emnlp.210.","DOI":"10.18653\/v1\/2021.findings-emnlp.210"},{"key":"9779_CR600","unstructured":"Westfall, C. (2023). Non-profit helpline shifts to chatbots, then shuts down rogue AI. Forbes. https:\/\/www.forbes.com\/sites\/chriswestfall\/2023\/05\/31\/non-profit-helpline-fires-staff-shifts-to-chatbot-solution\/?sh=67ec62e26dad."},{"key":"9779_CR60","unstructured":"Whitney, L. (2023). I Tried X\u2019s \u2018anti-Woke\u2019 Grok AI Chatbot. The Results Were the Opposite of What I Expected. ZDNET, 23 Dec. www.zdnet.com\/article\/i-tried-xs-anti-woke-grok-ai-chatbot-the-results-were-the-opposite-of-what-i-expected\/."},{"key":"9779_CR61","unstructured":"Wiggers, K. (2024). OpenAI\u2019s Chatbot Store Is Filling up with Spam. TechCrunch, 20 Mar. https:\/\/techcrunch.com\/2024\/03\/20\/openais-chatbot-store-is-filling-up-with-spam\/."},{"key":"9779_CR62","unstructured":"Williams, A., Miceli, M., & Gebru, T. (2022, October 13). The Exploited Labor behnd AI. Retrieved from NOEMA: https:\/\/www.noemamag.com\/the-exploited-labor-behind-artificial-intelligence\/."},{"key":"9779_CR63","unstructured":"Wu, Z., Qiu, L., Ross, A., Aky\u00fcrek, E., Chen, B., Wang, B., & Kim, Y. (2023). Reasoning or reciting? Exploring the capabilities and limitations of language models through counterfactual tasks. arXiv Preprint arXiv, 230702477."},{"key":"9779_CR64","doi-asserted-by":"publisher","unstructured":"Yiu, E., Kosoy, E., & Gopnik, A. (2023). Transmission Versus Truth, Imitation Versus Innovation: What children can do that large Language and Language-and-vision models cannot (yet). Perspectives on Psychological Science, 0(0). https:\/\/doi.org\/10.1177\/17456916231201401.","DOI":"10.1177\/17456916231201401"},{"key":"9779_CR65","unstructured":"Zhao, P., Zhang, H., Yu, Q., Wang, Z., Geng, Y., Fu, F., & Cui, B. (2024). Retrieval-Augmented Generation for AI-Generated Content: A Survey. arXiv Preprint arXiv, 2402, 19473."},{"key":"9779_CR66","doi-asserted-by":"publisher","unstructured":"Ziegler, D. M., Stiennon, N., Wu, J., Brown, T. B., Radford, A., Amodei, D. (2019). Fine-tuning language models from human preferences. https:\/\/doi.org\/10.48550\/arxiv.1909.08593.","DOI":"10.48550\/arxiv.1909.08593"}],"container-title":["Ethics and Information Technology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s10676-024-09779-1.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s10676-024-09779-1\/fulltext.html","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s10676-024-09779-1.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,6,20]],"date-time":"2024-06-20T13:28:55Z","timestamp":1718890135000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s10676-024-09779-1"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,5,31]]},"references-count":61,"journal-issue":{"issue":"2","published-print":{"date-parts":[[2024,6]]}},"alternative-id":["9779"],"URL":"https:\/\/doi.org\/10.1007\/s10676-024-09779-1","relation":{},"ISSN":["1388-1957","1572-8439"],"issn-type":[{"value":"1388-1957","type":"print"},{"value":"1572-8439","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,5,31]]},"assertion":[{"value":"31 May 2024","order":1,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"Not Applicable.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Ethical approval"}},{"value":"Not Applicable.","order":3,"name":"Ethics","group":{"name":"EthicsHeading","label":"Consent for publication"}},{"value":"Not Applicable.","order":4,"name":"Ethics","group":{"name":"EthicsHeading","label":"Consent for publication"}}],"article-number":"36"}}