Results 1 to 10 of about 1,011 (107)
Piccola — A Small Composition Language [PDF]
ECOOP Workshops, 1999 Piccola is a “small composition language” currently being developed within the Software Composition Group. The goal of Piccola is to support the flexible composition of applications from software components.
Oscar Nierstrasz
openaire +2 more sources
How Abilities in Large Language Models are Affected by Supervised Fine-tuning Data Composition [PDF]
Annual Meeting of the Association for Computational Linguistics, 2023 Large language models (LLMs) with enormous pre-training tokens and parameters emerge diverse abilities, including math reasoning, code generation, and instruction following. These abilities are further enhanced by supervised fine-tuning (SFT).
Guanting Dong +9 more
semanticscholar +1 more source
InternLM-XComposer: A Vision-Language Large Model for Advanced Text-image Comprehension and Composition [PDF]
arXiv.org, 2023 We propose InternLM-XComposer, a vision-language large model that enables advanced image-text comprehension and composition. The innovative nature of our model is highlighted by three appealing properties: 1) Interleaved Text-Image Composition: InternLM ...
Pan Zhang +18 more
semanticscholar +1 more source
LoraHub: Efficient Cross-Task Generalization via Dynamic LoRA Composition [PDF]
arXiv.org, 2023 Low-rank adaptations (LoRA) are often employed to fine-tune large language models (LLMs) for new tasks. This paper investigates LoRA composability for cross-task generalization and introduces LoraHub, a simple framework devised for the purposive assembly
Chengsong Huang +5 more
semanticscholar +1 more source
Sheared LLaMA: Accelerating Language Model Pre-training via Structured Pruning [PDF]
International Conference on Learning Representations, 2023 The popularity of LLaMA (Touvron et al., 2023a;b) and other recently emerged moderate-sized large language models (LLMs) highlights the potential of building smaller yet powerful LLMs.
Mengzhou Xia +3 more
semanticscholar +1 more source
XSTest: A Test Suite for Identifying Exaggerated Safety Behaviours in Large Language Models [PDF]
North American Chapter of the Association for Computational Linguistics, 2023 Without proper safeguards, large language models will readily follow malicious instructions and generate toxic content. This risk motivates safety efforts such as red-teaming and large-scale feedback learning, which aim to make models both helpful and ...
Paul Röttger +5 more
semanticscholar +1 more source
A Compositional Neural Architecture for Language [PDF]
Journal of Cognitive Neuroscience, 2018 AbstractHierarchical structure and compositionality imbue human language with unparalleled expressive power and set it apart from other perception–action systems. However, neither formal nor neurobiological models account for how these defining computational properties might arise in a physiological system.
Andrea E. Martin, Andrea E. Martin
openaire +6 more sources
Ignore Previous Prompt: Attack Techniques For Language Models [PDF]
arXiv.org, 2022 Transformer-based large language models (LLMs) provide a powerful foundation for natural language tasks in large-scale customer-facing applications. However, studies that explore their vulnerabilities emerging from malicious user interaction are scarce ...
Fábio Perez, I. Ribeiro
semanticscholar +1 more source
Modular language composition for the masses [PDF]
Proceedings of the 11th ACM SIGPLAN International Conference on Software Language Engineering, 2018 The goal of modular language development is to enable the definition of new languages as assemblies of pre-existing ones. Recent approaches in this area are plentiful but usually suffer from two main problems: either they do not support modular language composition both at the specification and implementation levels, or they require advanced knowledge ...
Leduc, Manuel +2 more
openaire +3 more sources
Crystal structure generation with autoregressive large language modeling [PDF]
Nature Communications, 2023 The generation of plausible crystal structures is often the first step in predicting the structure and properties of a material from its chemical composition.
Luis M. Antunes +2 more
semanticscholar +1 more source