Don't Drop Dropout: Optimizing Layer Sparsity for Efficient LLM Training and Inference Paper • 2609.05275 • Published 4 days ago • 11
deepseek-ai/DeepSeek-V4-Flash-Vision-Exp Image-Text-to-Text • 305B • Updated 6 days ago • 252k • • 785
Terminal-Universe: Turning Agent Trajectories into Scalable Terminal Environments Paper • 2609.04148 • Published 5 days ago • 276
SMELT: Scaling Laws for Compute-Matched MoE Looped Transformers Paper • 2609.01343 • Published 7 days ago • 98
On the Design of Qwen3.8-Next Architecture: Evaluation, Efficiency, and Training Stability Paper • 2608.30320 • Published 8 days ago • 53
DSpark: Confidence-Scheduled Speculative Decoding with Semi-Autoregressive Generation Paper • 2607.05147 • Published Jul 6 • 49
gittensor-model-hub/Qwen3.8-27B-NVFP4-RTX5090 Image-Text-to-Text • 15B • Updated about 4 hours ago • 583k • 151
Running Featured 454 Microduck Sandbox 🐤 454 Display a bold landing page with stylized title and button
Don't Retrain, Align: Adapting Autoregressive LMs to Diffusion LMs via Representation Alignment Paper • 2605.06885 • Published May 7 • 1
ParaTempo: Efficient Parallel Reasoning via Temporal Confidence Paper • 2608.16425 • Published 21 days ago • 40