Hallucinations or Attention Misdirection? The Path to Strategic Value Extraction in Business Using Large Language Models
Abstract: LLMs with transformer architecture have revolutionized the domain of text generation, setting unprecedented benchmarks. Despite their impressive capabilities, LLMs have been criticized for generating outcomes that deviate from factual accuracy or display logical inconsistencies, phenomena commonly referred to as hallucinations. This term, however, has often been misapplied to any results deviating from the instructor's expectations, which this paper defines as attention misdirection rather than true hallucinations. Understanding the distinction between hallucinations and attention misdirection becomes increasingly relevant in business contexts, where the ramifications of such errors can significantly impact the value extraction from these inherently pre-trained models. This paper highlights the best practices of the PGI, Persona, Grouping, and Intelligence, method, a strategic framework that achieved a remarkable error rate of only 3,15 percent across 4,000 responses generated by GPT in response to a real business challenge. It emphasizes that by equipping experimentation with knowledge, businesses can unlock opportunities for innovation through the use of these natively pre-trained models. This reinforces the notion that strategic application grounded in a skilled team can maximize the benefits of emergent technologies such as the LLMs.
- Communicating natural programs to humans and machines. Advances in Neural Information Processing Systems, 35:3731–3743, 2022.
- Muhammad Rayhan Bustam. The analysis of ambiguous structures through the structural ambiguity concept. Apollo Project-Jurnal Ilmiah Jurusan Sastra Inggris, 1(1), 2016.
- Monolingual or multilingual instruction tuning: Which makes a better alpaca. arXiv preprint arXiv:2309.08958, 2023.
- Attention mechanism, transformers, bert, and gpt: tutorial and survey. 2020.
- Hallucinations in large multilingual translation models. Transactions of the Association for Computational Linguistics, 11:1500–1517, 2023.
- Gisela Harras. Concepts in linguistics–concepts in natural language. In International Conference on Conceptual Structures, pages 13–26. Springer, 2000.
- Understanding morphology. Routledge, 2013.
- Aline Ioste. Transforming the Output of Generative Pre-trained Transformer: The Influence of the PGI Framework on Attention Dynamics. arXiv e-prints, art. arXiv:2308.13317, August 2023. doi: 10.48550/arXiv.2308.13317.
- Survey of hallucination in natural language generation. ACM Computing Surveys, 55(12):1–38, 2023.
- Rnn-lstm-gru based language transformation. Soft Computing, 23(24):13007–13024, 2019.
- adaptmllm: Fine-tuning multilingual language models on low-resource languages with integrated llm playgrounds. Information, 14(12):638, 2023.
- Geoffrey N Leech. Principles of pragmatics. Routledge, 2016.
- Paul B Lieberman. Hallucination: Philosophy and psychology, 2014.
- Data-efficient fine-tuning for llm-based recommendation. arXiv preprint arXiv:2401.17197, 2024.
- A survey on hallucination in large vision-language models. arXiv preprint arXiv:2402.00253, 2024.
- Sebastian Löbner. Understanding semantics. Routledge, 2014.
- Hallucination: Philosophy and psychology. MIT Press, 2013.
- Rnn and lstm. Deep Learning with Applications Using Python: Chatbots and Face, Object, and Speech Recognition With TensorFlow and Keras, pages 115–126, 2018.
- A comprehensive overview of large language models. arXiv preprint arXiv:2307.06435, 2023.
- Irwan Nursidi. Monolingual development in bilingual family: Exploring parental language policy within family domain. International Journal of multicultural and multireligious understanding, 6(3):219–224, 2019.
- Ian Pratt-hartmann. Computational complexity in natural language. The handbook of computational linguistics and natural language processing, pages 43–73, 2010.
- Diane Proudfoot. Anthropomorphism and ai turing. Artificial Intelligence, 175(5-6):950–957, 2011.
- Controlling hallucinations at word level in data-to-text generation. Data Mining and Knowledge Discovery, pages 1–37, 2022.
- Jennifer Rodd. Lexical ambiguity. Oxford handbook of psycholinguistics, pages 120–144, 2018.
- Rusdi Noor Rosa. Introduction to linguistics, 2013.
- The formal complexity of natural language, volume 33. Springer Science & Business Media, 2012.
- David A Schmidt. Programming language semantics. ACM Computing Surveys (CSUR), 28(1):265–267, 1996.
- John F Sowa. Lexical structures and conceptual structures. In Semantics and the Lexicon, pages 223–262. Springer, 1993.
- The psychology of human thought. CUP Archive, 1988.
- Maggie Tallerman. Understanding syntax. Routledge, 2019.
- Attention is all you need. Advances in neural information processing systems, 30, 2017.
- Cost-effective hyperparameter optimization for large language model generation inference. In International Conference on Automated Machine Learning, pages 21–1. PMLR, 2023.
- A survey on data selection for llm instruction tuning. arXiv preprint arXiv:2402.05123, 2024.
- On the cognitive process of human problem solving. Cognitive systems research, 11(1):81–92, 2010.
- A survey of resource-efficient llm and multimodal foundation models. arXiv preprint arXiv:2401.08092, 2024.
- Multimodal learning with transformers: A survey. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2023.
- To repeat or not to repeat: Insights from scaling llm under token-crisis. arXiv preprint arXiv:2305.13230, 2023.
- Investigating the catastrophic forgetting in multimodal large language models. arXiv preprint arXiv:2309.10313, 2023.
- Knowledge-enriched transformer for emotion detection in textual conversations. arXiv preprint arXiv:1909.10681, 2019.
- Toolqa: A dataset for llm question answering with external tools. Advances in Neural Information Processing Systems, 36, 2024.
Paper Prompts
Sign up for free to create and run prompts on this paper using GPT-5.
Top Community Prompts
Collections
Sign up for free to add this paper to one or more collections.