Phantom of Latent for Large Language and Vision Models

TL;DR

This paper introduces Phantom, an efficient large language and vision model family that enhances learning capacity by temporarily increasing latent dimensions, achieving high performance with fewer parameters.

Contribution

The paper proposes Phantom, a new LLVM architecture that improves efficiency by increasing latent dimensions during attention, combined with optimization techniques, outperforming larger models.

Findings

Phantom models outperform larger open- and closed-source LLVMs.

Temporary latent dimension increase enhances vision-language understanding.

Efficient models with fewer parameters achieve competitive performance.

Abstract

The success of visual instruction tuning has accelerated the development of large language and vision models (LLVMs). Following the scaling laws of instruction-tuned large language models (LLMs), LLVMs either have further increased their sizes, reaching 26B, 34B, and even 80B parameters. While this increase in model size has yielded significant performance gains, it demands substantially more hardware resources for both training and inference. Consequently, there naturally exists a strong need for efficient LLVMs that achieve the performance of larger models while being smaller in size. To achieve this need, we present a new efficient LLVM family with model sizes of 0.5B, 1.8B, 3.8B, and 7B parameters, Phantom, which significantly enhances learning capabilities within limited structures. By temporarily increasing the latent hidden dimension during multi-head self-attention (MHSA), we make LLVMs prepare to look and understand much more vision-language knowledge on the latent, without substantially increasing physical model sizes. To maximize its advantage, we introduce Phantom Optimization (PO) using both autoregressive supervised fine-tuning (SFT) and direct preference optimization (DPO)-like concept, which effectively follows correct answers while eliminating incorrect and ambiguous ones. Phantom outperforms numerous larger open- and closed-source LLVMs, positioning itself as a leading solution in the landscape of efficient LLVMs.