{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,6,21]],"date-time":"2025-06-21T18:40:08Z","timestamp":1750531208007,"version":"3.41.0"},"publisher-location":"Cham","reference-count":16,"publisher":"Springer Nature Switzerland","isbn-type":[{"value":"9783031961953","type":"print"},{"value":"9783031961960","type":"electronic"}],"license":[{"start":{"date-parts":[[2025,1,1]],"date-time":"2025-01-01T00:00:00Z","timestamp":1735689600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.springernature.com\/gp\/researchers\/text-and-data-mining"},{"start":{"date-parts":[[2025,1,1]],"date-time":"2025-01-01T00:00:00Z","timestamp":1735689600000},"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":[],"published-print":{"date-parts":[[2025]]},"DOI":"10.1007\/978-3-031-96196-0_19","type":"book-chapter","created":{"date-parts":[[2025,6,21]],"date-time":"2025-06-21T18:13:35Z","timestamp":1750529615000},"page":"254-266","update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Needle-in-the-Haystack Testing LLMs with\u00a0a\u00a0Complex Reasoning Task"],"prefix":"10.1007","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-9539-1627","authenticated-orcid":false,"given":"Thomas","family":"Schuster","sequence":"first","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0001-7347-5734","authenticated-orcid":false,"given":"Marian","family":"Lambert","sequence":"additional","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0009-0004-2018-5417","authenticated-orcid":false,"given":"Nico","family":"D\u00f6ring","sequence":"additional","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0009-0004-9937-7437","authenticated-orcid":false,"given":"Julius","family":"Tr\u00f6gele","sequence":"additional","affiliation":[]}],"member":"297","published-online":{"date-parts":[[2025,6,22]]},"reference":[{"key":"19_CR1","unstructured":"Baccianella, S.: JSON Repair - a python module to repair invalid JSON, commonly used to parse the output of LLMs, August 2024. https:\/\/github.com\/mangiucugna\/json_repair"},{"key":"19_CR2","doi-asserted-by":"crossref","unstructured":"Bai, Y., et al.: LongBench: a bilingual, multitask benchmark for long context understanding. In: Ku, L.W., Martins, A., Srikumar, V. (eds.) Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pp. 3119\u20133137. Association for Computational Linguistics, Bangkok, Thailand, August 2024. https:\/\/aclanthology.org\/2024.acl-long.172","DOI":"10.18653\/v1\/2024.acl-long.172"},{"key":"19_CR3","unstructured":"gkamradt: gkamradt\/LLMTest_needleinahaystack, October 2024, https:\/\/github.com\/gkamradt\/LLMTest_NeedleInAHaystack"},{"key":"19_CR4","doi-asserted-by":"publisher","unstructured":"Hsieh, C.P., et al.: RULER: what\u2019s the real context size of your long-context language models? August 2024. https:\/\/doi.org\/10.48550\/arXiv.2404.06654, https:\/\/arxiv.org\/abs\/2404.06654, arXiv:2404.06654 [cs]","DOI":"10.48550\/arXiv.2404.06654"},{"key":"19_CR5","unstructured":"HuggingFace: large language model text generation inference (2025). https:\/\/github.com\/huggingface\/text-generation-inference"},{"key":"19_CR6","doi-asserted-by":"publisher","unstructured":"Kuratov, Y., Bulatov, A., Anokhin, P., Sorokin, D., Sorokin, A., Burtsev, M.: In Search of needles in a 11m haystack: recurrent memory finds what LLMs miss, February 2024. https:\/\/doi.org\/10.48550\/arXiv.2402.10790, https:\/\/arxiv.org\/abs\/2402.10790, arXiv:2402.10790 [cs]","DOI":"10.48550\/arXiv.2402.10790"},{"key":"19_CR7","doi-asserted-by":"publisher","unstructured":"Laban, P., Fabbri, A.R., Xiong, C., Wu, C.S.: Summary of a haystack: a challenge to long-context LLMs and RAG systems (Jul 2024). https:\/\/doi.org\/10.48550\/arXiv.2407.01370, https:\/\/arxiv.org\/abs\/2407.01370, arXiv:2407.01370 [cs]","DOI":"10.48550\/arXiv.2407.01370"},{"key":"19_CR8","doi-asserted-by":"publisher","unstructured":"Lambert, M., D\u00f6ring, N., Schuster, T.: Evaluating large language models and prompt variants on the task of detecting cease and desist violations in german online product descriptions. DataMod 2023, LNCS 14618 (2025). https:\/\/doi.org\/10.1007\/978-3-031-87217-4_8","DOI":"10.1007\/978-3-031-87217-4_8"},{"key":"19_CR9","doi-asserted-by":"publisher","unstructured":"Lee, H., Sharma, S., Hu, B.: Bug in the code stack: can LLMs find bugs in large Python code stacks, June 2024. https:\/\/doi.org\/10.48550\/arXiv.2406.15325, https:\/\/arxiv.org\/abs\/2406.15325, arXiv:2406.15325 [cs]","DOI":"10.48550\/arXiv.2406.15325"},{"key":"19_CR10","doi-asserted-by":"publisher","unstructured":"Li, M., Zhang, S., Liu, Y., Chen, K.: NeedleBench: can LLMs do retrieval and reasoning in 1 million context window? July 2024. https:\/\/doi.org\/10.48550\/arXiv.2407.11963, https:\/\/arxiv.org\/abs\/2407.11963, arXiv:2407.11963 [cs]","DOI":"10.48550\/arXiv.2407.11963"},{"key":"19_CR11","doi-asserted-by":"publisher","unstructured":"Liu, N.F., et al.: Lost in the middle: how language models use long contexts, July 2023. https:\/\/doi.org\/10.48550\/arXiv.2307.03172, https:\/\/arxiv.org\/abs\/2307.03172, arXiv:2307.03172 [cs]","DOI":"10.48550\/arXiv.2307.03172"},{"key":"19_CR12","doi-asserted-by":"publisher","unstructured":"Machlab, D., Battle, R.: LLM in-context recall is prompt dependent, April 2024. https:\/\/doi.org\/10.48550\/arXiv.2404.08865, https:\/\/arxiv.org\/abs\/2404.08865, arXiv:2404.08865 [cs]","DOI":"10.48550\/arXiv.2404.08865"},{"key":"19_CR13","unstructured":"Meta: The Llama 3 Herd of Models | Research - AI at Meta, https:\/\/ai.meta.com\/research\/publications\/the-llama-3-herd-of-models\/"},{"key":"19_CR14","doi-asserted-by":"publisher","unstructured":"Waidelich, L., Lambert, M., Al-Washash, Z., Kroschwald, S., Schuster, T., D\u00f6ring, N.: Using large language models for the enforcement of consumer rights in Germany. In: Ma\u015blankowski, J., Marcinkowski, B., Rupino\u00a0da Cunha, P. (eds.) Digital Transformation, pp. 1\u201315. LNBIP, Springer, Cham (2023). https:\/\/doi.org\/10.1007\/978-3-031-43590-4_1","DOI":"10.1007\/978-3-031-43590-4_1"},{"key":"19_CR15","doi-asserted-by":"publisher","unstructured":"Wang, W., et al.: needle in a multimodal haystack, June 2024. https:\/\/doi.org\/10.48550\/arXiv.2406.07230, https:\/\/arxiv.org\/abs\/2406.07230, arXiv:2406.07230 [cs]","DOI":"10.48550\/arXiv.2406.07230"},{"key":"19_CR16","unstructured":"Zhao, W.X., et al.: A survey of large language models, June 2023. https:\/\/arxiv.org\/abs\/2303.18223, arXiv:2303.18223 [cs]"}],"container-title":["Communications in Computer and Information Science","Engineering Applications of Neural Networks"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/978-3-031-96196-0_19","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,6,21]],"date-time":"2025-06-21T18:13:40Z","timestamp":1750529620000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/978-3-031-96196-0_19"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025]]},"ISBN":["9783031961953","9783031961960"],"references-count":16,"URL":"https:\/\/doi.org\/10.1007\/978-3-031-96196-0_19","relation":{},"ISSN":["1865-0929","1865-0937"],"issn-type":[{"value":"1865-0929","type":"print"},{"value":"1865-0937","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025]]},"assertion":[{"value":"22 June 2025","order":1,"name":"first_online","label":"First Online","group":{"name":"ChapterHistory","label":"Chapter History"}},{"value":"EANN","order":1,"name":"conference_acronym","label":"Conference Acronym","group":{"name":"ConferenceInfo","label":"Conference Information"}},{"value":"International Conference on Engineering Applications of Neural Networks","order":2,"name":"conference_name","label":"Conference Name","group":{"name":"ConferenceInfo","label":"Conference Information"}},{"value":"Limassol","order":3,"name":"conference_city","label":"Conference City","group":{"name":"ConferenceInfo","label":"Conference Information"}},{"value":"Cyprus","order":4,"name":"conference_country","label":"Conference Country","group":{"name":"ConferenceInfo","label":"Conference Information"}},{"value":"2025","order":5,"name":"conference_year","label":"Conference Year","group":{"name":"ConferenceInfo","label":"Conference Information"}},{"value":"26 June 2025","order":7,"name":"conference_start_date","label":"Conference Start Date","group":{"name":"ConferenceInfo","label":"Conference Information"}},{"value":"29 June 2025","order":8,"name":"conference_end_date","label":"Conference End Date","group":{"name":"ConferenceInfo","label":"Conference Information"}},{"value":"eann2025a","order":10,"name":"conference_id","label":"Conference ID","group":{"name":"ConferenceInfo","label":"Conference Information"}},{"value":"https:\/\/eannconf.org\/2025\/","order":11,"name":"conference_url","label":"Conference URL","group":{"name":"ConferenceInfo","label":"Conference Information"}}]}}