Persistent Priors, Preserved Targets: A Stroop-Style Paradigm for Lexical Override
Not on the map. Why?
Abstract (arXiv)
Local definitions can assign a familiar word a temporary meaning while its usual associations remain useful elsewhere. We measure interference from those associations with a matched Stroop-style paradigm. A conflict prompt defines doctor as forest and compares forest with the familiar associate hospital. A neutral control replaces doctor with a semantically weak word in both the definition and query while keeping forest and hospital fixed. All 11 model-level means are positive. Aggregate means are also positive for all four conflict families and prompt formats. When no redefinition is present, a stronger preference for the familiar distractor predicts more interference in arbitrary-semantic, polysemy/entity, and domain-definition remappings, while the antonym slope is null. Separately, we patch neutral-control activations into antonym prompts in five 1B-2B models. Patching the defined word, the target word in the definition, and the later query word together restores almost all of the target-minus-distractor margin lost in conflict (normalized recovery R in [0.92,1.06]). Replacing only that target-word activation with a donor from another item reduces recovery in every tested case. Those donor patches also lower the distractor logit, while the contextual target falls much more than under the same-item patch that restores the margin.
Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 24 of its 43 references that are in the map. First posted on arXiv 25 May 2026.
Builds on
24 of 43 references.
References · 24
- 2022Locating and Editing Factual Associations in GPTcited by 2,609
- 2024Gemma 2: Improving Open Language Models at a Practical Sizecited by 1,928
- 2022Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 smallcited by 1,130
- 2021An Explanation of In-context Learning as Implicit Bayesian Inferencecited by 915
- 2023Large Language Models Can Be Easily Distracted by Irrelevant Contextcited by 820
- 2024OLMo: Accelerating the Science of Language Modelscited by 599
- 2023Larger language models do in-context learning differentlycited by 386
- 2021Entity-Based Knowledge Conflicts in Question Answeringcited by 349
- 2023Adaptive Chameleon or Stubborn Sloth: Revealing the Behavior of Large Language Models in Knowledge Conflictscited by 333
- 2023Towards Best Practices of Activation Patching in Language Models: Metrics and Methodscited by 328
- 2019What BERT is not: Lessons from a new suite of psycholinguistic diagnostics for language modelscited by 317
- 2024Knowledge Conflicts for LLMs: A Surveycited by 274
12 more
- 2019Negated and Misprimed Probes for Pretrained Language Models: Birds Can Talk, But Cannot Flycited by 240
- 2023What In-Context Learning "Learns" In-Context: Disentangling Task Recognition and Task Learningcited by 149
- 2023Inspecting and Editing Knowledge Representations in Language Modelscited by 140
- 2024Fine-Tuning Enhances Existing Mechanisms: A Case Study on Entity Trackingcited by 140
- 2024Cutting Off the Head Ends the Conflict: A Mechanism for Interpreting and Mitigating Knowledge Conflicts in Language Modelscited by 87
- 2024Steering Knowledge Selection Behaviours in LLMs via SAE-Based Representation Engineeringcited by 60
- 2020Exploring BERT's Sensitivity to Lexical Cues using Tests from Semantic Primingcited by 41
- 2024Competition of Mechanisms: Tracing How Language Models Handle Facts and Counterfactualscited by 40
- 2024Studying Large Language Model Behaviors Under Context-Memory Conflicts With Real Documentscited by 27
- 2025Task Matters: Knowledge Requirements Shape LLM Responses to Context-Memory Conflictcited by 15
- 2026When Models Ignore Definitions: Measuring Semantic Override Hallucinations in LLM Reasoningcited by 1
- 2026The Override Gap: A Magnitude Account of Knowledge Conflict Failure in Hypernetwork-Based Instant LLM Adaptationcited by 1
Built on by
Not cited yet.