Supporting Sensemaking of Large Language Model Outputs at Scale
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866916105354739712 |
|---|---|
| author | Gero, Katy Ilonka Swoopes, Chelse Gu, Ziwei Kummerfeld, Jonathan K. Glassman, Elena L. |
| author_facet | Gero, Katy Ilonka Swoopes, Chelse Gu, Ziwei Kummerfeld, Jonathan K. Glassman, Elena L. |
| contents | Large language models (LLMs) are capable of generating multiple responses to a single prompt, yet little effort has been expended to help end-users or system designers make use of this capability. In this paper, we explore how to present many LLM responses at once. We design five features, which include both pre-existing and novel methods for computing similarities and differences across textual documents, as well as how to render their outputs. We report on a controlled user study (n=24) and eight case studies evaluating these features and how they support users in different tasks. We find that the features support a wide variety of sensemaking tasks and even make tasks previously considered to be too difficult by our participants now tractable. Finally, we present design guidelines to inform future explorations of new LLM interfaces. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2401_13726 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Supporting Sensemaking of Large Language Model Outputs at Scale Gero, Katy Ilonka Swoopes, Chelse Gu, Ziwei Kummerfeld, Jonathan K. Glassman, Elena L. Human-Computer Interaction Machine Learning Large language models (LLMs) are capable of generating multiple responses to a single prompt, yet little effort has been expended to help end-users or system designers make use of this capability. In this paper, we explore how to present many LLM responses at once. We design five features, which include both pre-existing and novel methods for computing similarities and differences across textual documents, as well as how to render their outputs. We report on a controlled user study (n=24) and eight case studies evaluating these features and how they support users in different tasks. We find that the features support a wide variety of sensemaking tasks and even make tasks previously considered to be too difficult by our participants now tractable. Finally, we present design guidelines to inform future explorations of new LLM interfaces. |
| title | Supporting Sensemaking of Large Language Model Outputs at Scale |
| topic | Human-Computer Interaction Machine Learning |
| url | https://arxiv.org/abs/2401.13726 |