MegaScience: Pushing the Frontiers of Post-Training Datasets for Science Reasoning Paper โข 2507.16812 โข Published 7 days ago โข 46
PUSA V1.0: Surpassing Wan-I2V with $500 Training Cost by Vectorized Timestep Adaptation Paper โข 2507.16116 โข Published 7 days ago โข 9
NeuralOS: Towards Simulating Operating Systems via Neural Generative Models Paper โข 2507.08800 โข Published 18 days ago โข 74
The Devil behind the mask: An emergent safety vulnerability of Diffusion LLMs Paper โข 2507.11097 โข Published 14 days ago โข 56
NoHumansRequired: Autonomous High-Quality Image Editing Triplet Mining Paper โข 2507.14119 โข Published 11 days ago โข 49
SeC: Advancing Complex Video Object Segmentation via Progressive Concept Construction Paper โข 2507.15852 โข Published 8 days ago โข 36
GLM-4.1V-Thinking: Towards Versatile Multimodal Reasoning with Scalable Reinforcement Learning Paper โข 2507.01006 โข Published 28 days ago โข 200
Thinking with Images for Multimodal Reasoning: Foundations, Methods, and Future Frontiers Paper โข 2506.23918 โข Published 29 days ago โข 82
WebSailor: Navigating Super-human Reasoning for Web Agent Paper โข 2507.02592 โข Published 26 days ago โข 100
LongAnimation: Long Animation Generation with Dynamic Global-Local Memory Paper โข 2507.01945 โข Published 27 days ago โข 74
Seaweed-7B: Cost-Effective Training of Video Generation Foundation Model Paper โข 2504.08685 โข Published Apr 11 โข 129
view post Post 2895 I got rejected from llama4.So that means I can use quantinized model without following their TOS.Interesting. (JK) See translation 3 replies ยท ๐ 6 6 ๐ 2 2 ๐ 2 2 + Reply
view post Post 4765 Qwen 3 can launch very soon. ๐https://github.com/ggml-org/llama.cpp/pull/12828 See translation 3 replies ยท ๐ฅ 16 16 ๐ 9 9 โค๏ธ 8 8 + Reply
Running on Zero 820 820 MMAudio โ generating synchronized audio from video/text ๐ Generate audio from video or text prompts