Hugging Face Daily PapersJaihyun Lew, Mingi Jung, Minjun ParkSep 22, 20261 min readpaperadvanced

FoMo: Forking Moment in Generative Trajectory as a Perceptual Distance

Summary

FoMo introduces a fully automated method for generating perceptual distance labels between image pairs, eliminating the need for human annotation. It leverages the "forking moment" in a diffusion model's generative trajectory, where early divergence indicates large perceptual differences and late divergence indicates subtle ones.

FoMo automates perceptual distance labeling using diffusion model dynamics, bypassing expensive human annotation.
The "forking moment" in a diffusion model's generation process correlates with human perception of image differences.
Early forking implies coarse structural differences, while late forking implies fine detail differences.
FoMo generates pointwise labels, providing richer training data than traditional pairwise comparisons.

This paper is crucial for ML engineers and researchers developing or evaluating generative AI models, as it offers a scalable and objective way to measure image quality and perceptual similarity without relying on costly and inconsistent human judgments.

8/10

machine learning computer vision metrics image generation diffusion models

Related reading

FAMOS: Feed-Forward 3D Articulation Modeling from Sparse Observations arxiv.org
FAMOS is a feed-forward model that predicts 3D articulation, including movable-part segmentation and joint parameters, from sparse, unordered point clouds. It leverages multiple observations through a Multi-state Articulation Transformer and a novel observed articulation span objective, outperforming existing baselines.
Hugging Face Daily Papersarxiv.orgKevin Qu, Tao Sun, Massimiliano Viola19d1 minpaper
Why Do Video Diffusion Models Violate Physics? Unveiling the Flaws in Attention Mechanisms arxiv.org
The paper investigates why video diffusion models often break physical laws, pinpointing excessive spatial decay from Rotary Position Embedding (RoPE) as the culprit. By analyzing cross‑attention trajectories and self‑attention patterns, the authors identify specific attention heads that drive motion planning. They propose a lightweight fix: scaling RoPE frequency per denoising step, which empiri…
Hugging Face Daily Papersarxiv.orgYueyan Li, Haibo Wang, Caixia Yuan16d1 minpaper
LLaDA-UI: Bringing Block-wise Diffusion to Vision-Language GUI Agents arxiv.org
LLaDA‑UI is a 16.7 B‑parameter Mixture‑of‑Experts vision‑language model that uses block‑wise diffusion to generate GUI actions in parallel. It achieves state‑of‑the‑art performance on multiple GUI grounding and navigation benchmarks, beating larger Qwen models while keeping decoding latency low.
Hugging Face Daily Papersarxiv.orgZhangxuan Gu, Haoxing Chen, Qi Qin27d1 minpaper
Latent-Foresight: End-to-End Learning Predictable Representations for Latent World Models arxiv.org
Latent-Foresight is an end-to-end framework that jointly learns a latent tokenizer and a flow-based generative dynamics model for world modeling. It explicitly shapes representations for temporal predictability, outperforming two-stage baselines in future scene understanding tasks.
Hugging Face Daily Papersarxiv.orgEfstathios Karypidis, Spyros Gidaris, Nikos Komodakis5d1 minpaper
Scaffolding Minds: Optimizing Latent Visual Target Representations for Multimodal Reasoning arxiv.org
The paper pinpoints two weaknesses in latent‑reasoning pipelines—suboptimal visual token encoders and deterministic RL sampling—and proposes Scaffolding Minds, which trains a dedicated scaffolding encoder and a stochastic RL sampler (learning both mean and variance). This yields up to +19 points improvement on spatial planning tasks and +5.6 points average gain on nine visual reasoning benchmarks.
Hugging Face Daily Papersarxiv.orgHaoqiang Kang, Yinpeng Chen, Luyang Liu7d1 minpaper
Ego2Act: Evaluating Goal-Directed Manipulation in Egocentric Video Generation arxiv.org
Ego2Act introduces a new benchmark for evaluating goal-directed manipulation in egocentric video generation, featuring 2,640 videos across 110 real-world tasks. It reveals that current video generation models struggle with multi-step physical reasoning, often skipping steps and failing at fine-grained object manipulation and persistent world modeling.
Hugging Face Daily Papersarxiv.orgPatrick Amadeus Irawan, Iskandar Muda Rizky Parlambang, Rava Maulana5d1 minpaper

FoMo: Forking Moment in Generative Trajectory as a Perceptual Distance

Summary

Related reading

FAMOS: Feed-Forward 3D Articulation Modeling from Sparse Observationsarxiv.org

Why Do Video Diffusion Models Violate Physics? Unveiling the Flaws in Attention Mechanismsarxiv.org

LLaDA-UI: Bringing Block-wise Diffusion to Vision-Language GUI Agentsarxiv.org

Latent-Foresight: End-to-End Learning Predictable Representations for Latent World Modelsarxiv.org

Scaffolding Minds: Optimizing Latent Visual Target Representations for Multimodal Reasoningarxiv.org

Ego2Act: Evaluating Goal-Directed Manipulation in Egocentric Video Generationarxiv.org

FAMOS: Feed-Forward 3D Articulation Modeling from Sparse Observations arxiv.org

Why Do Video Diffusion Models Violate Physics? Unveiling the Flaws in Attention Mechanisms arxiv.org

LLaDA-UI: Bringing Block-wise Diffusion to Vision-Language GUI Agents arxiv.org

Latent-Foresight: End-to-End Learning Predictable Representations for Latent World Models arxiv.org

Scaffolding Minds: Optimizing Latent Visual Target Representations for Multimodal Reasoning arxiv.org

Ego2Act: Evaluating Goal-Directed Manipulation in Egocentric Video Generation arxiv.org