dorsal/arxiv
View SchemaIs Sentiment Banana-Shaped? Exploring the Geometry and Portability of Sentiment Concept Vectors
| Authors | Laurits Lyngbaek, Pascale Feldkamp, Yuri Bizzoni, Kristoffer L. Nielbo, Kenneth Enevoldsen |
|---|---|
| Categories | |
| ArXiv ID | 2601.07995vv1 |
| URL | https://arxiv.org/abs/2601.07995 |
| License | http://creativecommons.org/licenses/by-sa/4.0/ |
Abstract
Use cases of sentiment analysis in the humanities often require contextualized, continuous scores. Concept Vector Projections (CVP) offer a recent solution: by modeling sentiment as a direction in embedding space, they produce continuous, multilingual scores that align closely with human judgments. Yet the method's portability across domains and underlying assumptions remain underexplored. We evaluate CVP across genres, historical periods, languages, and affective dimensions, finding that concept vectors trained on one corpus transfer well to others with minimal performance loss. To understand the patterns of generalization, we further examine the linearity assumption underlying CVP. Our findings suggest that while CVP is a portable approach that effectively captures generalizable patterns, its linearity assumption is approximate, pointing to potential for further development.
{
"annotation_id": "c2e27386-7013-4eee-af4f-ab55dd663800",
"date_created": "2026-02-17T05:53:15.045000Z",
"date_modified": "2026-02-17T05:53:15.045000Z",
"file_hash": "6e084ed491a9d74ddb75e2800bd265d1940f0ed86abd420661eece371fd7cefd",
"private": false,
"record": {
"abstract": "Use cases of sentiment analysis in the humanities often require contextualized, continuous scores. Concept Vector Projections (CVP) offer a recent solution: by modeling sentiment as a direction in embedding space, they produce continuous, multilingual scores that align closely with human judgments. Yet the method\u0027s portability across domains and underlying assumptions remain underexplored. We evaluate CVP across genres, historical periods, languages, and affective dimensions, finding that concept vectors trained on one corpus transfer well to others with minimal performance loss. To understand the patterns of generalization, we further examine the linearity assumption underlying CVP. Our findings suggest that while CVP is a portable approach that effectively captures generalizable patterns, its linearity assumption is approximate, pointing to potential for further development.",
"arxiv_id": "2601.07995",
"authors": [
"Laurits Lyngbaek",
"Pascale Feldkamp",
"Yuri Bizzoni",
"Kristoffer L. Nielbo",
"Kenneth Enevoldsen"
],
"categories": [
"cs.CL"
],
"license": "http://creativecommons.org/licenses/by-sa/4.0/",
"title": "Is Sentiment Banana-Shaped? Exploring the Geometry and Portability of Sentiment Concept Vectors",
"url": "https://arxiv.org/abs/2601.07995",
"version": "v1"
},
"schema_id": "dorsal/arxiv",
"source": {
"execution_id": "d946ce0f-fa34-4fe2-916a-10d62b653ece",
"id": "arXiv Dataset",
"type": "Model",
"variant": "snapshot-2026-01-17",
"version": "0.1.0"
},
"user_id": 1000002
}