mlsanityData to 7 Oct 2026

Ask Again, Then Fail: Large Language Models' Vacillations in Judgment

arXiv 2310.02174 ↗First posted 3 Oct 2023cs.CL cs.AI cs.LGcited by 8

Not on the map. Why?

Abstract (arXiv)

We observe that current conversational language models often waver in their judgments when faced with follow-up questions, even if the original judgment was correct. This wavering presents a significant challenge for generating reliable responses and building user trust. To comprehensively assess this issue, we introduce a textscFollow-up Questioning Mechanism along with two metrics to quantify this inconsistency, confirming its widespread presence in current language models. To mitigate this issue, we explore various prompting strategies for closed-source models; moreover, we develop a training-based framework textscUnwavering-FQ that teaches language models to maintain their originally correct judgments through synthesized high-quality preference data. Our experimental results confirm the effectiveness of our framework and its ability to enhance the general capabilities of models.

Key facts. Cited by 8 papers in the map (data through 7 Oct 2026). Builds on 43 of its 79 references that are in the map. First posted on arXiv 3 Oct 2023.

Builds on

43 of 79 references.

References · 43

  1. 2020Language Models are Few-Shot Learnerscited by 32,199
  2. 2022Training language models to follow instructions with human feedbackcited by 15,712
  3. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 13,958
  4. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,302
  5. 2021Training Verifiers to Solve Math Word Problemscited by 9,743
  6. 2023Judging LLM-as-a-Judge with MT-Bench and Chatbot Arenacited by 8,393
  7. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 7,916
  8. 2020Measuring Massive Multitask Language Understandingcited by 7,768
  9. 2022PaLM: Scaling Language Modeling with Pathwayscited by 5,304
  10. 2022Large Language Models are Zero-Shot Reasonerscited by 4,323
  11. 2021On the Opportunities and Risks of Foundation Modelscited by 3,961
  12. 2022Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedbackcited by 3,634
31 more
  1. 2023QLoRA: Efficient Finetuning of Quantized LLMscited by 3,355
  2. 2023Universal and Transferable Adversarial Attacks on Aligned Language Modelscited by 2,914
  3. 2021Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processingcited by 2,222
  4. 2022Survey of Hallucination in Natural Language Generationcited by 2,056
  5. 2018CommonsenseQA: A Question Answering Challenge Targeting Commonsense Knowledgecited by 1,793
  6. 2022Language Models (Mostly) Know What They Knowcited by 1,783
  7. 2022Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?cited by 1,413
  8. 2021Calibrate Before Use: Improving Few-Shot Performance of Language Modelscited by 1,371
  9. 2023PaLM 2 Technical Reportcited by 1,154
  10. 2021Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order Sensitivitycited by 1,122
  11. 2023Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Promptingcited by 1,081
  12. 2021A General Language Assistant as a Laboratory for Alignmentcited by 1,023
  13. 2023Can LLMs Express Their Uncertainty? An Empirical Evaluation of Confidence Elicitation in LLMscited by 946
  14. 2023A Multitask, Multilingual, Multimodal Evaluation of ChatGPT on Reasoning, Hallucination, and Interactivitycited by 888
  15. 2021Are NLP Models really able to Solve Simple Math Word Problems?cited by 861
  16. 2023Large Language Models Can Be Easily Distracted by Irrelevant Contextcited by 820
  17. 2022Discovering Language Model Behaviors with Model-Written Evaluationscited by 805
  18. 2022Ignore Previous Prompt: Attack Techniques For Language Modelscited by 760
  19. 2021Did Aristotle Use a Laptop? A Question Answering Benchmark with Implicit Reasoning Strategiescited by 736
  20. 2022Teaching Models to Express Their Uncertainty in Wordscited by 731
  21. 2022Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learnedcited by 702
  22. 2016Solving General Arithmetic Word Problemscited by 353
  23. 2023Do Large Language Models Know What They Don't Know?cited by 238
  24. 2023Simple synthetic data reduces sycophancy in large language modelscited by 174
  25. 2022Learning by Distilling Contextcited by 128
  26. 2023Can ChatGPT Forecast Stock Price Movements? Return Predictability and Large Language Modelscited by 125
  27. 2023ReviewerGPT? An Exploratory Study on Using Large Language Models for Paper Reviewingcited by 118
  28. 2023Question Decomposition Improves the Faithfulness of Model-Generated Reasoningcited by 107
  29. 2023Is ChatGPT a Good Sentiment Analyzer? A Preliminary Studycited by 96
  30. 2023Matching Patients to Clinical Trials with Large Language Modelscited by 68
  31. 2023Beyond the Safeguards: Exploring the Security Risks of ChatGPTcited by 36

Built on by

Cited by year

Citing papers per year: 2024: 6; 2025: 2.

8 papers.

  1. 2024Understanding the Dark Side of LLMs' Intrinsic Self-Correctioncited by 42
  2. 2024BeHonest: Benchmarking Honesty in Large Language Modelscited by 20
  3. 2025Firm or Fickle? Evaluating Large Language Models Consistency in Sequential Interactionscited by 12
  4. 2024StructEval: Deepen and Broaden Large Language Model Assessment via Structured Evaluationcited by 9
  5. 2024A Survey on Responsible Generative AI: What to Generate and What Notcited by 8
  6. 2024ReflectDiffu:Reflect between Emotion-intent Contagion and Mimicry for Empathetic Response Generation via a RL-Diffusion Frameworkcited by 6
  7. 2025ATLAS: A High-Difficulty, Multidisciplinary Benchmark for Frontier Scientific Reasoningcited by 5
  8. 2024Why does in-context learning fail sometimes? Evaluating in-context learning on open and closed questionscited by 2