new

Get trending papers in your email inbox once a day!

Get trending papers in your email inbox!

Daily Papers

byAK and the research community

Dec 25

Submitted by

jt-zhang

TurboDiffusion: Accelerating Video Diffusion Models by 100-200 Times

UCBerkeley

University of California, Berkeley

Submitted by

AaronHuangWei

Learning to Reason in 4D: Dynamic Spatial Understanding for Vision Language Models

hkuhk

The University of Hong Kong

Submitted by

jwliu-cc

DreaMontage: Arbitrary Frame-Guided One-Shot Video Generation

ByteDance

1

Submitted by

CheeryLJH

T2AV-Compass: Towards Unified Evaluation for Text-to-Audio-Video Generation

NJU-LINK

Submitted by

yulunliu

Beyond Memorization: A Multi-Modal Ordinal Regression Benchmark to Expose Popularity Bias in Vision-Language Models

·
5 authors

Submitted by

MoonQiu

HiStream: Efficient High-Resolution Video Generation via Redundancy-Eliminated Streaming

Submitted by

taesiri

Nemotron 3 Nano: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoning

nvidia

Submitted by

taesiri

NVIDIA Nemotron 3: Efficient and Open Intelligence

nvidia

Submitted by

Malikeh1375

TokSuite: Measuring the Impact of Tokenizer Choice on Language Model Behavior

toksuite

Submitted by

feifeiobama

Learning from Next-Frame Prediction: Autoregressive Video Modeling Encodes Effective Representations

PekingUniversity

Peking University

Submitted by

Merlin-Hongru

From Word to World: Can Large Language Models be Implicit Text-based World Models?

EdinburghNLP

EdinburghNLP - Natural Language Processing Group at the University of Edinburgh

Submitted by

FutureMa

DramaBench: A Six-Dimensional Evaluation Framework for Drama Script Continuation

·
3 authors

Submitted by

bdqnghi

SWE-EVO: Benchmarking Coding Agents in Long-Horizon Software Evolution Scenarios

·
5 authors

Submitted by

taesiri

Streaming Video Instruction Tuning

·
5 authors

Submitted by

sinwang

Multi-hop Reasoning via Early Knowledge Alignment

·
7 authors

Submitted by

taesiri

LLM Swiss Round: Aggregating Multi-Benchmark Performance via Competitive Swiss-System Dynamics

ByteDance-Seed

Submitted by

xiao-qi

PhononBench:A Large-Scale Phonon-Based Benchmark for Dynamical Stability in Crystal Generation

RUC

Renmin University of China