GEAR: Guided End-to-End AutoRegression for Image Synthesis
GEAR trains a vector-quantized tokenizer and autoregressive generator jointly end-to-end using representation alignment, overcoming non-differentiability issues through a dual read…
Hugging Face · Daily Papers
·Bin Lin, Zheyuan Liu
·
·▲ 28 upvotes
Este artigo está em destaque na seleção diária de papers do Hugging Face, curada pela comunidade de pesquisa em IA.
Autores: Bin Lin, Zheyuan Liu, Chenguo Lin, Sixiang Chen, Yunyang Ge, Yunlong Lin
- 28 upvotes da comunidade
- Temas: vector-quantized, autoregressive, representation alignment, straight-through estimator, codebook assignment, next-token prediction
Resumo
Resumo original (em inglês), extraído do paper:
GEAR trains a vector-quantized tokenizer and autoregressive generator jointly end-to-end using representation alignment, overcoming non-differentiability issues through a dual read-out approach that improves convergence speed and feature quality.