AI Steerability 360: A Toolkit for Steering Large Language Models
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866910046018863104 |
|---|---|
| author | Miehling, Erik Ramamurthy, Karthikeyan Natesan Venkateswaran, Praveen Ko, Irene Dognin, Pierre Singh, Moninder Pedapati, Tejaswini Balakrishnan, Avinash Riemer, Matthew Wei, Dennis Vejsbjerg, Inge Daly, Elizabeth M. Varshney, Kush R. |
| author_facet | Miehling, Erik Ramamurthy, Karthikeyan Natesan Venkateswaran, Praveen Ko, Irene Dognin, Pierre Singh, Moninder Pedapati, Tejaswini Balakrishnan, Avinash Riemer, Matthew Wei, Dennis Vejsbjerg, Inge Daly, Elizabeth M. Varshney, Kush R. |
| contents | The AI Steerability 360 toolkit is an extensible, open-source Python library for steering LLMs. Steering abstractions are designed around four model control surfaces: input (modification of the prompt), structural (modification of the model's weights or architecture), state (modification of the model's activations and attentions), and output (modification of the decoding or generation process). Steering methods exert control on the model through a common interface, termed a steering pipeline, which additionally allows for the composition of multiple steering methods. Comprehensive evaluation and comparison of steering methods/pipelines is facilitated by use case classes (for defining tasks) and a benchmark class (for performance comparison on a given task). The functionality provided by the toolkit significantly lowers the barrier to developing and comprehensively evaluating steering methods. The toolkit is Hugging Face native and is released under an Apache 2.0 license at https://github.com/IBM/AISteer360. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_07837 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | AI Steerability 360: A Toolkit for Steering Large Language Models Miehling, Erik Ramamurthy, Karthikeyan Natesan Venkateswaran, Praveen Ko, Irene Dognin, Pierre Singh, Moninder Pedapati, Tejaswini Balakrishnan, Avinash Riemer, Matthew Wei, Dennis Vejsbjerg, Inge Daly, Elizabeth M. Varshney, Kush R. Computation and Language Artificial Intelligence The AI Steerability 360 toolkit is an extensible, open-source Python library for steering LLMs. Steering abstractions are designed around four model control surfaces: input (modification of the prompt), structural (modification of the model's weights or architecture), state (modification of the model's activations and attentions), and output (modification of the decoding or generation process). Steering methods exert control on the model through a common interface, termed a steering pipeline, which additionally allows for the composition of multiple steering methods. Comprehensive evaluation and comparison of steering methods/pipelines is facilitated by use case classes (for defining tasks) and a benchmark class (for performance comparison on a given task). The functionality provided by the toolkit significantly lowers the barrier to developing and comprehensively evaluating steering methods. The toolkit is Hugging Face native and is released under an Apache 2.0 license at https://github.com/IBM/AISteer360. |
| title | AI Steerability 360: A Toolkit for Steering Large Language Models |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2603.07837 |