mlsanityData to 14 Oct 2026

How Prompts Move Language Model Behavior: Frames, Salience, and Construal as Semantic Control

arXiv 2512.12688 ↗First posted 14 Dec 2025cs.LG cs.AI cs.CLcited by 0

Not on the map. Why?

Abstract (arXiv)

Prompt engineering is widely used to shape large language model behavior, yet it is often treated as a practical heuristic rather than as a form of natural-language control. This paper develops a cognitive-semantic account in which prompts function as semantic conditions on how a fixed model interprets inputs, foregrounds information, and structures tasks. We formalize this account through three notions -- frame activation, salience control, and construal selection -- and study them in natural language inference, claim verification, and multi-hop question answering. Across these settings, prompts produce measurable changes in label judgments, evidence use, and answer-support organization, showing that prompt effects differ not only in magnitude but also in semantic direction. The paper therefore reframes prompting as the analysis of how instructions move model behavior, rather than only whether they improve performance.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 59 of its 86 references that are in the map. First posted on arXiv 14 Dec 2025.

Builds on

59 of 86 references.

References · 59

  1. 2020Language Models are Few-Shot Learnerscited by 32,419
  2. 2022Training language models to follow instructions with human feedbackcited by 15,971
  3. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,459
  4. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,297
  5. 2022Self-Consistency Improves Chain of Thought Reasoning in Language Modelscited by 5,708
  6. 2022Large Language Models are Zero-Shot Reasonerscited by 4,374
  7. 2021Prefix-Tuning: Optimizing Continuous Prompts for Generationcited by 3,915
  8. 2020Dense Passage Retrieval for Open-Domain Question Answeringcited by 3,824
  9. 2021The Power of Scale for Parameter-Efficient Prompt Tuningcited by 3,795
  10. 2021Finetuned Language Models Are Zero-Shot Learnerscited by 3,780
  11. 2018HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answeringcited by 3,754
  12. 2023Lost in the Middle: How Language Models Use Long Contextscited by 3,455
38 more
  1. 2022Scaling Instruction-Finetuned Language Modelscited by 3,024
  2. 2021Pre-train, Prompt, and Predict: A Systematic Survey of Prompting Methods in Natural Language Processingcited by 2,234
  3. 2020REALM: Retrieval-Augmented Language Model Pre-Trainingcited by 2,141
  4. 2021Multitask Prompted Training Enables Zero-Shot Task Generalizationcited by 1,470
  5. 2022Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?cited by 1,434
  6. 2021Calibrate Before Use: Improving Few-Shot Performance of Language Modelscited by 1,392
  7. 2018FEVER: a large-scale dataset for Fact Extraction and VERificationcited by 1,375
  8. 2023FActScore: Fine-grained Atomic Evaluation of Factual Precision in Long Form Text Generationcited by 1,282
  9. 2020Making Pre-trained Language Models Better Few-shot Learnerscited by 1,256
  10. 2021Fantastically Ordered Prompts and Where to Find Them: Overcoming Few-Shot Prompt Order Sensitivitycited by 1,135
  11. 2019Attention is not Explanationcited by 929
  12. 2021MuSiQue: Multihop Questions via Single-hop Question Compositioncited by 916
  13. 2022Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Taskscited by 868
  14. 2023Large Language Models Can Be Easily Distracted by Irrelevant Contextcited by 841
  15. 2023Quantifying Language Models' Sensitivity to Spurious Features in Prompt Design or: How I learned to start worrying about prompt formattingcited by 817
  16. 2021Did Aristotle Use a Laptop? A Question Answering Benchmark with Implicit Reasoning Strategiescited by 743
  17. 2021Cross-Task Generalization via Natural Language Crowdsourcing Instructionscited by 699
  18. 2021Learning To Retrieve Prompts for In-Context Learningcited by 648
  19. 2023Enabling Large Language Models to Generate Text with Citationscited by 595
  20. 2020KILT: a Benchmark for Knowledge Intensive Language Taskscited by 573
  21. 2019ERASER: A Benchmark to Evaluate Rationalized NLP Modelscited by 530
  22. 2018The Web as a Knowledge-base for Answering Complex Questionscited by 410
  23. 2019Is Attention Interpretable?cited by 389
  24. 2021Do Prompt-Based Models Really Understand the Meaning of their Prompts?cited by 326
  25. 2016Rationalizing Neural Predictionscited by 281
  26. 2021Surface Form Competition: Why the Highest Probability Answer Isn't Always Rightcited by 263
  27. 2021Measuring Attribution in Natural Language Generation Modelscited by 256
  28. 2018Learning to Explain: An Information-Theoretic Perspective on Model Interpretationcited by 248
  29. 2022Active Example Selection for In-Context Learningcited by 196
  30. 2017Constructing Datasets for Multi-hop Reading Comprehension Across Documentscited by 190
  31. 2020Open Question Answering over Tables and Textcited by 163
  32. 2022Self-Adaptive In-Context Learning: An Information Compression Perspective for In-Context Example Selection and Orderingcited by 161
  33. 2020A Diagnostic Study of Explainability Techniques for Text Classificationcited by 159
  34. 2023Large Language Models Are Latent Variable Models: Explaining and Finding Good Demonstrations for In-Context Learningcited by 149
  35. 2023Compositional Exemplars for In-context Learningcited by 144
  36. 2020Break It Down: A Question Understanding Benchmarkcited by 131
  37. 2019Interpretable Neural Predictions with Differentiable Binary Variablescited by 126
  38. 2019Multi-hop Reading Comprehension through Question Decomposition and Rescoringcited by 119

50 most cited of 59 All 59 (JSON)

Built on by

Not cited yet.