ALTo: Adaptive-Length Tokenizer for Autoregressive Mask Generation

TL;DR

ALTo introduces an adaptive-length tokenizer for autoregressive mask generation, enabling multimodal models to allocate attention dynamically based on visual complexity, improving segmentation performance and efficiency.

Contribution

The paper presents a novel adaptive-length tokenizer and a seamless integration method into multimodal large language models, enhancing mask quality and computational efficiency.

Findings

Achieves state-of-the-art segmentation results

Demonstrates improved efficiency with adaptive token costs

Provides a flexible framework for mask quality and efficiency trade-offs

Abstract

While humans effortlessly draw visual objects and shapes by adaptively allocating attention based on their complexity, existing multimodal large language models (MLLMs) remain constrained by rigid token representations. Bridging this gap, we propose ALTo, an adaptive length tokenizer for autoregressive mask generation. To achieve this, a novel token length predictor is designed, along with a length regularization term and a differentiable token chunking strategy. We further build ALToLLM that seamlessly integrates ALTo into MLLM. Preferences on the trade-offs between mask quality and efficiency is implemented by group relative policy optimization (GRPO). Experiments demonstrate that ALToLLM achieves state-of-the-art performance with adaptive token cost on popular segmentation benchmarks. Code and models are released at https://github.com/yayafengzi/ALToLLM.