Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
102 tokens/sec
GPT-4o
59 tokens/sec
Gemini 2.5 Pro Pro
43 tokens/sec
o3 Pro
6 tokens/sec
GPT-4.1 Pro
50 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

RELIC: Investigating Large Language Model Responses using Self-Consistency (2311.16842v2)

Published 28 Nov 2023 in cs.HC and cs.CL

Abstract: LLMs are notorious for blending fact with fiction and generating non-factual content, known as hallucinations. To address this challenge, we propose an interactive system that helps users gain insight into the reliability of the generated text. Our approach is based on the idea that the self-consistency of multiple samples generated by the same LLM relates to its confidence in individual claims in the generated texts. Using this idea, we design RELIC, an interactive system that enables users to investigate and verify semantic-level variations in multiple long-form responses. This allows users to recognize potentially inaccurate information in the generated text and make necessary corrections. From a user study with ten participants, we demonstrate that our approach helps users better verify the reliability of the generated text. We further summarize the design implications and lessons learned from this research for future studies of reliable human-LLM interactions.

PDF HTML Abstract
Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Bookmark Chat Bubble Oval Streamline Icon: https://streamlinehq.com Chat (Pro)
Definition Search Book Streamline Icon: https://streamlinehq.com
References (51)
  1. Ask Me Anything: A simple strategy for prompting language models. In The Eleventh International Conference on Learning Representations.
  2. Uncertainty as a form of transparency: Measuring, communicating, and using uncertainty. In Proceedings of the 2021 AAAI/ACM Conference on AI, Ethics, and Society. 401–413.
  3. Mapping the Design Space of Human-AI Interaction in Text Summarization. In Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. 431–455.
  4. Avishek Choudhury and Hamid Shamszare. 2023. Investigating the Impact of User Trust on the Adoption and Use of ChatGPT: Survey Analysis. Journal of Medical Internet Research 25 (2023), 0.
  5. Beyond Text Generation: Supporting Writers with Continuous Automatic Text Summaries. In Proceedings of the 35th Annual ACM Symposium on User Interface Software and Technology (Bend, OR, USA) (UIST ’22). Article 98, 13 pages.
  6. Zijian Ding and Joel Chan. 2023. Mapping the Design Space of Interactions in Human-AI Text Co-creation Tasks. arXiv:2303.06430
  7. Is ChatGPT a Highly Fluent Grammatical Error Correction System? A Comprehensive Evaluation. arXiv:2304.01746
  8. Bridging the gap: A survey on integrating (human) feedback for natural language generation. arXiv:2305.00955
  9. Unsupervised quality estimation for neural machine translation. Transactions of the Association for Computational Linguistics 8 (2020), 539–555.
  10. Exploring the placement and design of word-scale visualizations. IEEE Transactions on Visualization and Computer Graphics 20, 12 (2014), 2291–2300.
  11. CRITIC: Large language models can self-correct with tool-interactive critiquing. arXiv:2305.11738
  12. Creative Writing with an AI-Powered Writing Assistant: Perspectives from Professional Writers. arXiv:2211.05030
  13. Survey of Hallucination in Natural Language Generation. ACM Comput. Surv. 55, 12, Article 248 (2023), 38 pages.
  14. Promptmaker: Prompt-based prototyping with large language models. In CHI Conference on Human Factors in Computing Systems Extended Abstracts. 1–8.
  15. Graphologue: Exploring Large Language Model Responses with Interactive Diagrams. arXiv:2305.11473
  16. How can we know when language models know? on the calibration of language models for question answering. Transactions of the Association for Computational Linguistics.
  17. Language Models (Mostly) Know What They Know. arXiv:2207.05221
  18. Large Language Models Struggle to Learn Long-Tail Knowledge. Proceedings of the 40th International Conference on Machine Learning (2023).
  19. Semantic Uncertainty: Linguistic Invariances for Uncertainty Estimation in Natural Language Generation. In The Eleventh International Conference on Learning Representations.
  20. Coauthor: Designing a human-ai collaborative writing dataset for exploring language model capabilities. In Proceedings of the 2022 CHI conference on human factors in computing systems. 1–19.
  21. Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks. arXiv:2005.11401
  22. Faithfulness in Natural Language Generation: A Systematic Survey of Analysis, Evaluation and Optimization Methods. arXiv:2203.05227
  23. Q Vera Liao and Jennifer Wortman Vaughan. 2023. AI Transparency in the Age of LLMs: A Human-Centered Research Roadmap. arXiv:2306.01941
  24. Zachary C Lipton. 2018. The mythos of model interpretability: In machine learning, the concept of interpretability is both important and slippery. Queue 16, 3 (2018), 31–57.
  25. SelfCheckGPT: Zero-Resource Black-Box Hallucination Detection for Generative Large Language Models. arXiv:2303.08896
  26. FActScore: Fine-grained Atomic Evaluation of Factual Precision in Long Form Text Generation. arXiv:2305.14251
  27. Self-contradictory Hallucinations of Large Language Models: Evaluation, Detection and Mitigation. arXiv:2305.15852
  28. Anglekindling: Supporting journalistic angle ideation with large language models. In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems. 1–16.
  29. Improving Wikipedia Verifiability with AI.
  30. ”Why should i trust you?” Explaining the predictions of any classifier. In Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining. 1135–1144.
  31. Visual comparison of language model adaptation. IEEE Transactions on Visualization and Computer Graphics 29, 1 (2022), 1178–1188.
  32. Rita Sevastjanova and Mennatallah El-Assady. 2022. Beware the rationalization trap! when language model explainability diverges from our mental models of language. arXiv:2207.06897
  33. LMFingerprints: Visual explanations of language model embedding spaces through layerwise contextualization scores. In Computer Graphics Forum. Wiley Online Library, 295–307.
  34. explAIner: A visual analytics framework for interactive and explainable machine learning. IEEE Transactions on Visualization and computer graphics 26, 1 (2019), 1064–1074.
  35. Seq2seq-vis: A visual debugging tool for sequence-to-sequence models. IEEE transactions on visualization and computer graphics 25, 1 (2018), 353–363.
  36. Lstmvis: A tool for visual analysis of hidden state dynamics in recurrent neural networks. IEEE transactions on visualization and computer graphics 24, 1 (2017), 667–676.
  37. Interactive and visual prompt engineering for ad-hoc task adaptation with large language models. IEEE transactions on visualization and computer graphics 29, 1 (2022), 1146–1156.
  38. Sensecape: Enabling Multilevel Exploration and Sensemaking with Large Language Models. arXiv:2305.11483
  39. The language interpretability tool: Extensible, interactive visualizations and analysis for NLP models. arXiv:2008.05122
  40. FEVER: A Large-scale Dataset for Fact Extraction and VERification. In Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long Papers). 809–819. https://doi.org/10.18653/v1/N18-1074
  41. Do NLP Models Know Numbers? Probing Numeracy in Embeddings. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processin.
  42. Designing theory-driven user-centric explainable AI. In Proceedings of the 2019 CHI conference on human factors in computing systems. 1–15.
  43. Chain-of-thought prompting elicits reasoning in large language models. Advances in Neural Information Processing Systems 35 (2022), 24824–24837.
  44. Promptchainer: Chaining large language model prompts through visual programming. In CHI Conference on Human Factors in Computing Systems Extended Abstracts. 1–10.
  45. AI chains: Transparent and controllable human-AI interaction by chaining large language model prompts. In Proceedings of the 2022 CHI conference on human factors in computing systems. 1–22.
  46. AI as an Active Writer: Interaction strategies with generated text in human-AI collaborative fiction writing. In Joint Proceedings of the ACM IUI Workshops, Vol. 10. CEUR-WS Team.
  47. It is AI’s Turn to Ask Humans a Question: Question-Answer Pair Generation for Children’s Story Books. arXiv:2109.03423
  48. Wordcraft: Story Writing With Large Language Models. In 27th International Conference on Intelligent User Interfaces (Helsinki, Finland) (IUI ’22). 841–852. https://doi.org/10.1145/3490099.3511105
  49. Why Johnny can’t prompt: How non-AI experts try (and fail) to design LLM prompts. In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems. 1–21.
  50. Michael J.Q. Zhang and Eunsol Choi. 2021. Situatedqa: Incorporating extra-linguistic contexts into QA. Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processin (2021).
  51. Effect of confidence and explanation on accuracy and trust calibration in AI-assisted decision making. In FAT* ’20: Conference on Fairness, Accountability, and Transparency Barcelona, Spain, January 27-30, 2020, Mireille Hildebrandt, Carlos Castillo, L. Elisa Celis, Salvatore Ruggieri, Linnet Taylor, and Gabriela Zanfir-Fortuna (Eds.). 295–305.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (6)
  1. Furui Cheng (10 papers)
  2. Vilém Zouhar (41 papers)
  3. Simran Arora (64 papers)
  4. Mrinmaya Sachan (124 papers)
  5. Hendrik Strobelt (43 papers)
  6. Mennatallah El-Assady (54 papers)
Citations (8)
View on Semantic Scholar
X Twitter Logo Streamline Icon: https://streamlinehq.com

Tweets