-
Quo Vadis, World Modeling?
score 9
入选 HF Daily Papers；HF 热度: 33 upvotes (+4)；有代码实现
-
Knowledge-Geometry Decoupling: Refreshable Pretrained Transfer for Streaming Recommendation
score 10
入选 HF Daily Papers；HF 热度: 47 upvotes (+4)；有代码实现；关键词(1): production
-
PAST-Bench: Benchmarking the Foundations of Recursive Self-Improvement in Personal Agents
score 9
入选 HF Daily Papers；HF 热度: 31 upvotes (+4)；有代码实现
-
When Teachers Mislead: Spurious-Signal-Aware On-Policy Distillation
score 10
入选 HF Daily Papers；HF 热度: 20 upvotes (+4)；有代码实现；关键词(4): lightweight, distillation, reasoning, vision-language
-
Video-DeepResearch: Towards the Next-Generation Multimodal Deepresearch Agent
score 8
入选 HF Daily Papers；HF 热度: 48 upvotes (+4)；关键词(2): fine-tuning, GRPO
-
LLaDA MoE v2: Scaling Mixture-of-Experts Diffusion Language Models
score 8
入选 HF Daily Papers；HF 热度: 29 upvotes (+4)；关键词(6): scaling, fine-tuning, pretraining, MoE, coding
-
Any-OPD: Heterogeneous On-Policy Distillation for Flow-Matching Models via Representation-Space Bridging
score 8
入选 HF Daily Papers；HF 热度: 26 upvotes (+4)；关键词(1): distillation
-
ContinualSkillBench: Can LLM Agents Truly Evolve Their Capabilities?
score 8
入选 HF Daily Papers；HF 热度: 13 upvotes (+3)；有代码实现
-
Hunyuan3D-Buffalo 1.0: A Unified Multimodal Model for Scalable 3D Generation, Understanding, and Editing
score 10
机构: Tencent；入选 HF Daily Papers；HF 热度: 79 upvotes (+4)
-
ARCHead: Activation-Metric Residual Correction for Large Language Model Output Heads
score 8
入选 HF Daily Papers；HF 热度: 6 upvotes (+2)；有代码实现；关键词(2): quantization, throughput