mlsanityData to 14 Oct 2026

Lexical Redefinition Reveals Context-Sensitive Instruction Following

arXiv 2610.16952 ↗First posted 12 Oct 2026cs.CLcited by 0

Not on the map. Why?

Abstract (arXiv)

How do language models reconcile conflicting contextual evidence within a prompt? We study this question using lexical redefinition, where an explicit instruction specifies how a word should be interpreted and a subsequent sentence provides additional evidence through its usage. Across 870 word pairs, we find that lexical redefinitions are not applied as fixed rules. Instead, the behaviour depends systematically on the relationship between the explicit instruction and the intervening sentence: the contextualised usage can preserve, weaken, or completely reverse the instruction. These results suggest that the interpretation of an explicit instruction remains sensitive to subsequent contextual evidence. Lexical redefinition therefore provides a controlled behavioural paradigm for studying how language models integrate multiple sources of information within a prompt.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 16 of its 32 references that are in the map. First posted on arXiv 12 Oct 2026.

Builds on

16 of 32 references.

References · 16

  1. 2020Language Models are Few-Shot Learnerscited by 32,419
  2. 2023Not what you've signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injectioncited by 1,443
  3. 2022Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?cited by 1,434
  4. 2021An Explanation of In-context Learning as Implicit Bayesian Inferencecited by 938
  5. 2023Large Language Models Can Be Easily Distracted by Irrelevant Contextcited by 841
  6. 2021Cross-Task Generalization via Natural Language Crowdsourcing Instructionscited by 699
  7. 2023Larger language models do in-context learning differentlycited by 394
  8. 2019What BERT is not: Lessons from a new suite of psycholinguistic diagnostics for language modelscited by 320
  9. 2023In-Context Learning Creates Task Vectorscited by 312
  10. 2024Knowledge Conflicts for LLMs: A Surveycited by 283
  11. 2019Negated and Misprimed Probes for Pretrained Language Models: Birds Can Talk, But Cannot Flycited by 243
  12. 2018Diachronic word embeddings and semantic shifts: a surveycited by 92
4 more
  1. 2022Event knowledge in large language models: the gap between the impossible and the unlikelycited by 45
  2. 2024Rectifying Demonstration Shortcut in In-Context Learningcited by 9
  3. 2024Semantic Change Characterization with LLMs using Rhetoricscited by 5
  4. 2025Semantic Anchors in In-Context Learning: Why Small LLMs Cannot Flip Their Labelscited by 2

Built on by

Not cited yet.