Han-Bit Kang

hbkang's activity

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

upvoted a paper 2 days ago

Chinese-Jev: Bringing System One Model to Chinese-Language Tasks

Paper • 2609.36965 • Published 3 days ago • 15

upvoted 2 papers 15 days ago

The Other Half of the Memory Wall: Serving 35B MoEs from SSD with Trained Routing Prediction

Paper • 2609.18063 • Published 16 days ago • 19

VC-Attention: Value Smoothing and Softmax Casting for Low-bit Attention

Paper • 2609.15810 • Published 18 days ago • 51

upvoted a paper 19 days ago

DFlash: Block Diffusion for Flash Speculative Decoding

Paper • 2602.06036 • Published Feb 5 • 100

upvoted a paper 20 days ago

Graph Machine: Towards Better Pretraining via Edges

Paper • 2609.02881 • Published about 1 month ago • 7

upvoted a paper 22 days ago

OracleZoom: On-Policy Self-Distillation Inspired Reference-Constrained Recursive Image Super Resolution

Paper • 2609.06490 • Published 26 days ago • 9

upvoted 2 papers 23 days ago

Unlocking Lossless Speedups in LLMs via Discrete Diffusion

Paper • 2609.04010 • Published 29 days ago • 114

Marigold V2: Revisiting Diffusion Transformers for Monocular Depth Estimation

Paper • 2609.08084 • Published 24 days ago • 72

upvoted a paper 24 days ago

ENEAS: Embedding-guided Neural Ensemble for Adaptive Segmentation

Paper • 2609.03756 • Published 29 days ago • 25

upvoted a paper 25 days ago

Puro-2B: Poor Lab's Qwen2-1.5B Trained on RTX 5090 within $5090

Paper • 2608.27370 • Published Aug 27 • 40

upvoted a paper 29 days ago

A Glance Is All You Need: Single-Pass Fine-Grained Image Captioning with SimLoss

Paper • 2609.00591 • Published Sep 1 • 18

upvoted 2 papers about 1 month ago

Keep-or-Drop? Adaptive Tokenizer for Compact Video Representation

Paper • 2608.24293 • Published Aug 25 • 14

Sliding-window beats linear attention

Paper • 2608.28444 • Published Aug 28 • 18

upvoted 2 papers about 2 months ago

FactorJEPA: Factorizing Monolithic Futures into Layout-Agent-Interaction Channels for Crowded and Chaotic Global South Urban Worlds

Paper • 2608.01049 • Published Aug 2 • 13

LeapTalk: Breaking the Latency-Quality Trade-off in Talking Head Generation

Paper • 2608.00079 • Published Jul 29 • 19

upvoted 3 papers 2 months ago

Memory for Large Language Models

Paper • 2607.25380 • Published Jul 28 • 13

P-MTP: Efficient Document Parsing via Multi-Token Prediction with Progressive Depth Scaling

Paper • 2606.24447 • Published Jun 23 • 1

HPD-Parsing: Hierarchical Parallel Document Parsing

Paper • 2607.18839 • Published Jul 21 • 14

upvoted 2 papers 3 months ago

Gemma 4 Technical Report

Paper • 2607.02770 • Published Jul 2 • 84

AgenticSTS: A Bounded-Memory Testbed for Long-Horizon LLM Agents

Paper • 2607.02255 • Published Jul 2 • 69