J C

dark-pen's activity

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

upvoted a paper about 10 hours ago

From RGB Generation to Dense Field Readout: Pixel-Space Dense Prediction with Text-to-Image Models

Paper • 2607.06553 • Published 5 days ago • 10

upvoted a paper about 15 hours ago

Splat and Distill: Augmenting Teachers with Feed-Forward 3D Reconstruction For 3D-Aware Distillation

Paper • 2602.06032 • Published Feb 5 • 2

upvoted 8 papers 1 day ago

Boxer: Robust Lifting of Open-World 2D Bounding Boxes to 3D

Paper • 2604.05212 • Published Apr 6 • 4

UniCorrn: Unified Correspondence Transformer Across 2D and 3D

Paper • 2605.04044 • Published May 5 • 1

Efficient Part-level 3D Object Generation via Dual Volume Packing

Paper • 2506.09980 • Published Jun 11, 2025 • 8

SceneEval: Evaluating Semantic Coherence in Text-Conditioned 3D Indoor Scene Synthesis

Paper • 2503.14756 • Published Mar 18, 2025 • 2

HSM: Hierarchical Scene Motifs for Multi-Scale Indoor Scene Generation

Paper • 2503.16848 • Published Dec 5, 2025 • 3

ReVSI: Rebuilding Visual Spatial Intelligence Evaluation for Accurate Assessment of VLM 3D Reasoning

Paper • 2604.24300 • Published Apr 27 • 69

GenRecon: Bridging Generative Priors for Multi-View 3D Scene Reconstruction

Paper • 2605.23888 • Published May 22 • 16

PEFT-Arena: Understanding Parameter-Efficient Finetuning from a Stability-Plasticity Perspective

Paper • 2605.28819 • Published May 27 • 9

upvoted 3 papers 2 days ago

CADFit: Precise Mesh-to-CAD Program Generation with Hybrid Optimization

Paper • 2605.01171 • Published Jun 7 • 1

CADBench: A Multimodal Benchmark for AI-Assisted CAD Program Generation

Paper • 2605.10873 • Published May 11 • 2

CADReasoner: Iterative Program Editing for CAD Reverse Engineering

Paper • 2603.29847 • Published Feb 18 • 3

upvoted a collection 2 days ago

Qwen3.5-Claude-Fable-5

Collection

Our series of Qwen3.5 finetunes on Claude-Fable-5 outputs • 2 items • Updated 25 days ago • 7

upvoted 3 papers 3 days ago

Img2CAD: Reverse Engineering 3D CAD Models from Images through VLM-Assisted Conditional Factorization

Paper • 2408.01437 • Published Jul 19, 2024 • 2

WildCity: A Real-World City-Scale Testbed for Rendering, Simulation, and Spatial Intelligence

Paper • 2607.06838 • Published 7 days ago • 13

ARDY: Autoregressive Diffusion with Hybrid Representation for Interactive Human Motion Generation

Paper • 2607.08741 • Published 5 days ago • 6

upvoted a paper 4 days ago

DINO-SAE: DINO Spherical Autoencoder for High-Fidelity Image Reconstruction and Generation

Paper • 2601.22904 • Published Jan 30 • 16

upvoted an article 5 days ago

view article

Article

Distillation in 2026 (so far): which frontier models use it and how

• 

sergiopaniego

•

6 days ago

• 15

upvoted a paper 5 days ago

Random Long-Context Access for Mamba via Hardware-aligned Hierarchical Sparse Attention

Paper • 2504.16795 • Published Apr 23, 2025 • 1