Over-Refusal and Representation Subspaces: A Mechanistic Analysis of Task-Conditioned Refusal in Aligned LLMs
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918528630652928 |
|---|---|
| author | Maskey, Utsav Dras, Mark Naseem, Usman |
| author_facet | Maskey, Utsav Dras, Mark Naseem, Usman |
| contents | Aligned language models that are trained to refuse harmful requests also exhibit over-refusal: they decline safe instructions that seemingly resemble harmful instructions. A natural approach is to ablate the global refusal direction, steering the hidden-state vectors away or towards the harmful-refusal examples, but this corrects over-refusal only incidentally while disrupting the broader refusal mechanism. In this work, we analyse the representational geometry of both refusal types to understand why this happens. We show that harmful-refusal directions are task-agnostic and can be captured by a single global vector, whereas over-refusal directions are task-dependent: they reside within the benign task-representation clusters, vary across tasks, and span a higher-dimensional subspace. Linear probing suggests that the two refusal types are representationally distinct from the early transformer layers. These findings provide a mechanistic explanation of why global direction ablation alone cannot address over-refusal, and establish that task-specific geometric interventions are necessary. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_27518 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Over-Refusal and Representation Subspaces: A Mechanistic Analysis of Task-Conditioned Refusal in Aligned LLMs Maskey, Utsav Dras, Mark Naseem, Usman Computation and Language Aligned language models that are trained to refuse harmful requests also exhibit over-refusal: they decline safe instructions that seemingly resemble harmful instructions. A natural approach is to ablate the global refusal direction, steering the hidden-state vectors away or towards the harmful-refusal examples, but this corrects over-refusal only incidentally while disrupting the broader refusal mechanism. In this work, we analyse the representational geometry of both refusal types to understand why this happens. We show that harmful-refusal directions are task-agnostic and can be captured by a single global vector, whereas over-refusal directions are task-dependent: they reside within the benign task-representation clusters, vary across tasks, and span a higher-dimensional subspace. Linear probing suggests that the two refusal types are representationally distinct from the early transformer layers. These findings provide a mechanistic explanation of why global direction ablation alone cannot address over-refusal, and establish that task-specific geometric interventions are necessary. |
| title | Over-Refusal and Representation Subspaces: A Mechanistic Analysis of Task-Conditioned Refusal in Aligned LLMs |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2603.27518 |