Research highlights
The “Super Weight:” How Even a Single Parameter can Determine a Large Language Model’s Behavior
August 21, 2025
A recent paper from Apple researchers, “The Super Weight in Large Language Models,” reveals that an extremely small subset of parameters in LLMs (in some cases, a single parameter) can exert a disproportionate influence on an LLM’s overall functionality (see Figure 1). This work highlights the critical role of these “super weights” and their corresponding “super activations,”…
FastVLM: Efficient Vision Encoding for Vision Language Models
July 23, 2025research area Computer Vision
Vision Language Models (VLMs) enable visual understanding alongside textual inputs. They are typically built by passing visual tokens from a pretrained vision encoder to a pretrained Large Language Model (LLM) through a projection layer. By leveraging the rich visual representations of the vision encoder and the world knowledge and reasoning capabilities of the LLM, VLMs can be useful for a wide range of applications, including accessibility…
Recent publications
Events
International Conference on Computer Vision (ICCV) 2025
October 13, 2025research area Computer Vision
Apple is presenting new work at the biennial International Conference on Computer Vision (ICCV), which takes place in person from October 19 to 23, in Honolulu, Hawai’i. The conference alternates each year with the European Conference on Computer Vision (ECCV), and focuses on important topics the field of computer vision.
Apple Workshop on Natural Language and Interactive Systems 2025
September 23, 2025
Natural language processing (NLP) remains one of the most quickly evolving fields in AI, as new research continues to rapidly advance large language models (LLMs), systems for speech recognition and generation, language agents, and more. This technology is essential to many of today’s AI experiences, including Apple Intelligence and Siri, and fundamental research in NLP will be foundational to future AI.