Papers1055

SALAD: Achieve High-Sparsity Attention via Efficient Linear Attention Tuning for Video Diffusion Transformer

Tongcheng Fang, Hanling Zhang et al.Jan 23arXiv

Videos are made of very long lists of tokens, and regular attention looks at every pair of tokens, which is slow and expensive.

#SALAD#sparse attention#linear attention

Endless Terminals: Scaling RL Environments for Terminal Agents

Intermediate

Kanishk Gandhi, Shivam Garg et al.Jan 23arXiv

Endless Terminals is an automatic factory that builds thousands of realistic, checkable computer-terminal tasks so AI agents can practice and improve with reinforcement learning.

#reinforcement learning#PPO#terminal agents

Memory-V2V: Augmenting Video-to-Video Diffusion Models with Memory

Intermediate

Dohun Lee, Chun-Hao Paul Huang et al.Jan 22arXiv

Memory-V2V teaches video editing AIs to remember what they already changed so new edits stay consistent with old ones.

#multi-turn video editing#video-to-video diffusion#explicit memory

GameTalk: Training LLMs for Strategic Conversation

Intermediate

Victor Conchello Vendrell, Max Ruiz Luyten et al.Jan 22arXiv

Large language models usually get judged one message at a time, but many real tasks need smart planning across a whole conversation.

#strategic conversation#reinforcement learning for LLMs#multi-turn dialogue

A Mechanistic View on Video Generation as World Models: State and Dynamics

Intermediate

Luozhou Wang, Zhifei Chen et al.Jan 22arXiv

This paper says modern video generators are starting to act like tiny "world simulators," not just pretty video painters.

#world models#video generation#state representation

Scaling Text-to-Image Diffusion Transformers with Representation Autoencoders

Intermediate

Shengbang Tong, Boyang Zheng et al.Jan 22arXiv

Before this work, most text-to-image models used VAEs (small, squished image codes) and struggled with slow training and overfitting on high-quality fine-tuning sets.

#Representation Autoencoder#RAE#Variational Autoencoder

360Anything: Geometry-Free Lifting of Images and Videos to 360°

Intermediate

Ziyi Wu, Daniel Watson et al.Jan 22arXiv

This paper shows how to turn any normal photo or video into a seamless 360° panorama without needing the camera’s settings like field of view or tilt.

#360 panorama generation#equirectangular projection#diffusion transformer

Learning to Discover at Test Time

Intermediate

Mert Yuksekgonul, Daniel Koceja et al.Jan 22arXiv

This paper shows how to keep training a language model while it is solving one hard, real problem, so it can discover a single, truly great answer instead of many average ones.

#test-time training#reinforcement learning#entropic objective

Cosmos Policy: Fine-Tuning Video Models for Visuomotor Control and Planning

Intermediate

Moo Jin Kim, Yihuai Gao et al.Jan 22arXiv

Cosmos Policy teaches robots to act by fine-tuning a powerful video model in just one training stage, without changing the model’s architecture.

#video diffusion#robot policy learning#visuomotor control

ActionMesh: Animated 3D Mesh Generation with Temporal 3D Diffusion

Intermediate

Remy Sabathier, David Novotny et al.Jan 22arXiv

ActionMesh is a fast, feed-forward AI that turns videos, images + text, text alone, or a given 3D model into an animated 3D mesh.

#ActionMesh#temporal 3D diffusion#animated 3D mesh

Rethinking Composed Image Retrieval Evaluation: A Fine-Grained Benchmark from Image Editing

Intermediate

Tingyu Song, Yanzhao Zhang et al.Jan 22arXiv

This paper introduces EDIR, a new and much more detailed test for Composed Image Retrieval (CIR), where you search for a target image using a starting image plus a short text change.

#Composed Image Retrieval#EDIR#fine-grained benchmark

SAMTok: Representing Any Mask with Two Words

Intermediate

Yikang Zhou, Tao Zhang et al.Jan 22arXiv

SAMTok turns any object’s mask in an image into just two special “words” so language models can handle pixels like they handle text.

#SAMTok#mask tokenizer#residual vector quantization

41 42 43 44 45