SeongWan Kim

idgmatrix's activity

• All
• Models
• Datasets
• Spaces
• Buckets
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

upvoted 9 papers 4 days ago

PN-Mix: Pair-Norm Guided Mixed-Precision Key Cache Quantization for Efficient Long-Output LLM Inference

Paper • 2610.19605 • Published 6 days ago • 4

Speculative Forcing for Efficient Autoregressive Video Generation

Paper • 2610.22347 • Published 5 days ago • 80

Unlocking Structured Pruning for Reasoning Language Models via Attention-Preserving MLP Trimming

Paper • 2610.19121 • Published 7 days ago • 5

Joint Subspace Sparse Attention for Efficient Long-Context Decoding

Paper • 2610.19026 • Published 7 days ago • 14

HumanoidToolBench: Benchmarking Humanoid Tool Use from Selection to Mobile Execution

Paper • 2610.02089 • Published 19 days ago • 34

KV4-RL: Unlocking Faster Asynchronous Rollouts for RLVR with INT4 KV Cache Quantization

Paper • 2610.14448 • Published 10 days ago • 44

MixLID: Token-Level Language Identification in Mixed-Language Text

Paper • 2610.16601 • Published 8 days ago • 4

Omni-Modal LLMs Know More Than They Say

Paper • 2610.16135 • Published 8 days ago • 10

Latent World Action Model for GUI Agents

Paper • 2610.14228 • Published 11 days ago • 27

upvoted 3 papers 6 days ago

Fish Audio Drama 3 Technical Report

Paper • 2610.18010 • Published 7 days ago • 33

Learning How to Remember: Memory Modality Selection for GUI Agents

Paper • 2610.17928 • Published 7 days ago • 47

Evaluating LLM Creativity through Long-Tail Performance

Paper • 2610.12998 • Published 11 days ago • 12

upvoted 3 papers 8 days ago

Continually Post-training Small Models via Synthetic Self-distillation

Paper • 2610.13113 • Published 11 days ago • 45

One Block, Multiple Depths: Recurrent Vision Transformers with Depth-Programmed Experts

Paper • 2610.12448 • Published 12 days ago • 11

Multi-Agent Egocentric World Model with Fine-Grained Embodied Interaction

Paper • 2610.12299 • Published 12 days ago • 72

upvoted 5 papers 12 days ago

DLoop: Looped Speculative Decoding

Paper • 2610.07659 • Published 14 days ago • 15

Improving Proactive AI Assistance with Hierarchical Procedural Understanding

Paper • 2610.06505 • Published 14 days ago • 13

On-Policy Distillation with Negative-Policy Rollouts

Paper • 2610.07874 • Published 14 days ago • 21

Tetris3D: 3D Scene Generation With Objects That Fit Together

Paper • 2610.10539 • Published 13 days ago • 49

RunningTab: Direct Workspace Interaction with Environment-Side Tabs

Paper • 2610.10444 • Published 13 days ago • 42