Pretrained Hybrids with MAD Skills

Nicholas Roberts, Samuel Guo, Zhiqi Gao, Satya Sai Srinath Namburi GNVV, Sonia Cromp, Chengjun Wu, Chengyu Duan, Frederic Sala

Keywords: hybrid architectures, large language models, transformers, state space models, model merging, neural architecture search, mechanistic interpretability

TL;DR: We develop a framework for creating pretrained hybrid models from existing pretrained models.

Abstract: While Transformers underpin modern large language models (LMs), there is a growing list of alternative architectures with new capabilities, promises, and tradeoffs. This makes choosing the right LM architecture challenging. Recently-proposed *hybrid architectures* seek a best-of-all-worlds approach that reaps the benefits of all architectures. Hybrid design is difficult for two reasons: it requires manual expert-driven search, and new hybrids must be trained from scratch. We propose **Manticore,** a framework that addresses these challenges. Manticore *automates the design of hybrid architectures* while reusing pretrained models to create *pretrained* hybrids. Our approach augments ideas from differentiable Neural Architecture Search (NAS) by incorporating simple projectors that translate features between pretrained blocks from different architectures. We then fine-tune hybrids that combine pretrained models from different architecture families---such as the GPT series and Mamba---end-to-end. With Manticore, we enable LM selection without training multiple models, the construction of pretrained hybrids from existing pretrained models, and the ability to *program* pretrained hybrids to have certain capabilities. Manticore hybrids outperform existing hybrids, achieve strong performance on Long Range Arena (LRA) tasks, and can improve on pretrained transformers and state space models.

Submission Number: 12