Don't Think of the White Bear: Ironic Negation in Transformer Models Under Cognitive Load
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909905482416128 |
|---|---|
| author | Mann, Logan Saxena, Nayan Tandon, Sarah Sun, Chenhao Toteja, Savar Zhu, Kevin |
| author_facet | Mann, Logan Saxena, Nayan Tandon, Sarah Sun, Chenhao Toteja, Savar Zhu, Kevin |
| contents | Negation instructions such as 'do not mention $X$' can paradoxically increase the accessibility of $X$ in human thought, a phenomenon known as ironic rebound. Large language models (LLMs) face the same challenge: suppressing a concept requires internally activating it, which may prime rebound instead of avoidance. We investigated this tension with two experiments. \textbf{(1) Load \& content}: after a negation instruction, we vary distractor text (semantic, syntactic, repetition) and measure rebound strength. \textbf{(2) Polarity separation}: We test whether models distinguish neutral from negative framings of the same concept and whether this separation predicts rebound persistence. Results show that rebound consistently arises immediately after negation and intensifies with longer or semantic distractors, while repetition supports suppression. Stronger polarity separation correlates with more persistent rebound. Together, these findings, complemented by a circuit tracing analysis that identifies sparse middle-layer attention heads amplifying forbidden tokens while early layers suppress, link cognitive predictions of ironic rebound with mechanistic insights into long-context interference. To support future work, we release ReboundBench, a dataset of $5,000$ systematically varied negation prompts designed to probe rebound in LLMs. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_12381 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Don't Think of the White Bear: Ironic Negation in Transformer Models Under Cognitive Load Mann, Logan Saxena, Nayan Tandon, Sarah Sun, Chenhao Toteja, Savar Zhu, Kevin Computation and Language Artificial Intelligence Negation instructions such as 'do not mention $X$' can paradoxically increase the accessibility of $X$ in human thought, a phenomenon known as ironic rebound. Large language models (LLMs) face the same challenge: suppressing a concept requires internally activating it, which may prime rebound instead of avoidance. We investigated this tension with two experiments. \textbf{(1) Load \& content}: after a negation instruction, we vary distractor text (semantic, syntactic, repetition) and measure rebound strength. \textbf{(2) Polarity separation}: We test whether models distinguish neutral from negative framings of the same concept and whether this separation predicts rebound persistence. Results show that rebound consistently arises immediately after negation and intensifies with longer or semantic distractors, while repetition supports suppression. Stronger polarity separation correlates with more persistent rebound. Together, these findings, complemented by a circuit tracing analysis that identifies sparse middle-layer attention heads amplifying forbidden tokens while early layers suppress, link cognitive predictions of ironic rebound with mechanistic insights into long-context interference. To support future work, we release ReboundBench, a dataset of $5,000$ systematically varied negation prompts designed to probe rebound in LLMs. |
| title | Don't Think of the White Bear: Ironic Negation in Transformer Models Under Cognitive Load |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2511.12381 |