跳到正文
HuggingFace Daily Papers(社區熱門論文)·· 2026-10-04AI 評分44

LiFT: Loop Flow Transformers

LiFT: Loop Flow Transformers

AI 導讀

LiFT 是一系列迴圈式生成模型,透過重複套用共享的 Diffusion Transformer(DiT)核心擴展計算,並可在無需重新訓練或修改的情況下,超越訓練深度進行迴圈推理。在 ImageNet 256x256 測試中,LiFT-L/2 的 FID 比密集型 DiT-XL/2 基線低 3.34,參數量約少 60%,訓練 FLOPs 少 32%,推理 FLOPs 少 52%。

正文

Published on Oct 4

Authors:

,

,

,

Abstract

We introduce Loop Flow Transformers (LiFT), a family of looped generative models that scales computation by repeatedly applying a shared Diffusion Transformer (DiT) core, with only light changes to the standard architecture. Rather than asking every recurrent step for the final prediction, LiFT trains each step with a single regression target: a point on a straight path from the model's initial estimate to the flow-matching target. Because we index these targets by a continuous depth coordinate, a trained model can loop far beyond its training depth with no retraining, early exits, or other modifications. In our experiments, these longer rollouts improve generation, so inference computation can grow without adding parameters. On ImageNet at 256x256, LiFT-L/2 achieves an FID 3.34 points lower than our dense DiT-XL/2 baseline while using approximately 60% fewer parameters, 32% fewer training FLOPs, and 52% fewer inference FLOPs.

View arXiv page View PDF Add to collection

Get this paper in your agent:

hf papers read 2610.05538

Don't have the latest CLI?

curl -LsSf https://hf.co/cli/install.sh | bash

Models citing this paper 0

No model linking this paper

Cite arxiv.org/abs/2610.05538 in a model README.md to link it from this page.

Datasets citing this paper 0

No dataset linking this paper

Cite arxiv.org/abs/2610.05538 in a dataset README.md to link it from this page.

Spaces citing this paper 0

No Space linking this paper

Cite arxiv.org/abs/2610.05538 in a Space README.md to link it from this page.

Collections including this paper 1

來源:HuggingFace Daily Papers(社區熱門論文) · huggingface.co