Abstract:Large language models have made significant progress in the past few years. However, they are either generic {\it or} field specific, splitting the community into different groups. In this paper, we unify these large language models into a larger map, where the generic {\it and} specific models are linked together and can improve each other, based on the user personal input and information from the internet. The idea of linking several large language models together is inspired by the functionality of human brain. The specific regions on the brain cortex are specific for certain low level functionality. And these regions can jointly work together to achieve more complex high level functionality. Such behavior on human brain cortex sheds the light to design the multilevel large language models that contain global level, field level and user level models. The user level models run on local machines to achieve efficient response and protect the user's privacy. Such multilevel models reduce some redundancy and perform better than the single level models. The proposed multilevel idea can be applied in various applications, such as natural language processing, computer vision tasks, professional assistant, business and healthcare.

What problem does this paper attempt to address?

The main problems that this paper attempts to solve are the limitations of existing large - language models (LLMs), specifically including: 1. **High consumption of computing resources**: Traditional large - language models require a large amount of computing resources (such as thousands of GPUs) for training and deployment, which is a huge obstacle for many researchers and organizations. 2. **Model staticness**: Existing large - language models are static and cannot be dynamically adjusted and evolved according to user input, thus limiting their performance improvement in specific fields. 3. **Privacy issues**: Since these models require a large amount of data for training, they may capture sensitive information, leading to the risk of user privacy leakage. 4. **Lack of domain - specificity**: Although general - purpose large - language models are applicable to a wide range of fields, their performance in specific fields is not excellent. To solve these problems, the paper proposes a multilevel large - scale language model framework, aiming to improve existing models in the following ways: - **Global Level**: It contains models with a very large number of parameters. These models are trained based on all available data sets, have extensive knowledge but are not focused on any specific domain. They are mainly used to update and improve domain - specific models. - **Field Level**: Models optimized for specific domains (such as poetry, medical diagnosis, programming, etc.). These models obtain knowledge from the global model and are further fine - tuned to improve performance in specific domains. - **User Level**: Small - scale models running on local machines can be customized according to users' personal preferences, protecting user privacy while providing efficient responses. In addition, the paper also proposes an economic ecosystem. Through blockchain technology, distributed computing is realized, the threshold for developing and using large - language models is lowered, and positive interactions between users and developers are encouraged, forming a sustainable development ecosystem. In conclusion, the goal of this paper is to solve the deficiencies of existing large - language models in terms of computing cost, dynamic adaptability, privacy protection, and domain - specificity by constructing a multilevel language model framework, thereby promoting the development of natural - language processing technology.

Multilevel Large Language Models for Everyone

PLMM: Personal Large Language Models on Mobile Devices

How do Large Language Models Handle Multilingualism?

A Survey of Large Language Models

Multilingual Large Language Models and Curse of Multilinguality

Multilingual Large Language Model: A Survey of Resources, Taxonomy and Frontiers

Large Language Model as a Universal Clinical Multi-task Decoder

Language-Specific Neurons: The Key to Multilingual Capabilities in Large Language Models

Converging to a Lingua Franca: Evolution of Linguistic Regions and Semantics Alignment in Multilingual Large Language Models

Demystifying Large Language Models for Medicine: A Primer

Multilingual Large Language Models: A Systematic Survey

Large Language Models in the Medical Field: Principles and Applications

Survey of different Large Language Model Architectures: Trends, Benchmarks, and Challenges

Large Language Models for Scientific Synthesis, Inference and Explanation

Contextual Feature Extraction Hierarchies Converge in Large Language Models and the Brain

Research on the Application and Optimization Strategies of Deep Learning in Large Language Models

MindLLM: Pre-training Lightweight Large Language Model from Scratch, Evaluations and Domain Applications

Multimodal Large Language Models: A Survey

A Comprehensive Survey of Large Language Models and Multimodal Large Language Models in Medicine

Are Large Language Models the New Interface for Data Pipelines?

Responsible Multilingual Large Language Models: A Survey of Development, Applications, and Societal Impact