Byeongho Heo

bhheo's activity

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

upvoted a paper 19 days ago

Revisiting Complete Reasoning Traces for Post-Training

Paper • 2609.07103 • Published 23 days ago • 23

upvoted a paper 20 days ago

Eliciting Weak-to-Strong Generalization with On-Policy Reverse Distillation

Paper • 2609.08798 • Published 22 days ago • 84

upvoted a paper 21 days ago

Beneath the Surface of Chains-of-Thought: A Mechanistic Interpretation of Reasoning Operations in LLMs

Paper • 2609.04753 • Published 26 days ago • 16

upvoted a paper 26 days ago

Verification-Aware Training for Speculative Decoding

Paper • 2608.30135 • Published 30 days ago • 11

upvoted a paper about 2 months ago

On-Policy Delta Distillation for Multilingual Math Reasoning

Paper • 2608.05802 • Published Aug 6 • 33

upvoted 3 papers 2 months ago

MuCo: Multi-turn Contrastive Learning for Multimodal Embedding Model

Paper • 2602.06393 • Published Feb 6 • 5

Weak-to-Strong Generalization via Direct On-Policy Distillation

Paper • 2607.05394 • Published Jul 8 • 144

On-Policy Delta Distillation

Paper • 2607.15161 • Published Jul 16 • 38

upvoted a paper 4 months ago

Retrieve, Don't Retrain: Extending Vision Language Action Models to New Tasks at Test Time

Paper • 2606.15631 • Published Jun 14 • 17

upvoted a paper 5 months ago

RLDX-1 Technical Report

Paper • 2605.03269 • Published May 5 • 128

upvoted a collection 6 months ago

MuCo

Collection

MuCo: Multi-turn Contrastive Learning for Multimodal Embedding Model [CVPR 2026] • 4 items • Updated Apr 13 • 2

upvoted a paper 7 months ago

Grounding World Simulation Models in a Real-World Metropolis

Paper • 2603.15583 • Published Mar 16 • 154

upvoted 3 papers 11 months ago

Exploring Conditions for Diffusion models in Robotic Control

Paper • 2510.15510 • Published Oct 17, 2025 • 40

Map the Flow: Revealing Hidden Pathways of Information in VideoLLMs

Paper • 2510.13251 • Published Oct 15, 2025 • 14

RL makes MLLMs see better than SFT

Paper • 2510.16333 • Published Oct 18, 2025 • 49

upvoted a paper about 1 year ago

Token Bottleneck: One Token to Remember Dynamics

Paper • 2507.06543 • Published Jul 9, 2025 • 20

upvoted 2 collections over 1 year ago

HyperCLOVA X SEED

Collection

HyperCLOVA X SEED is NAVER's lightweight open-source lineup with a strong focus on Korean language performance • 6 items • Updated Dec 24, 2025 • 43

ProLIP

Collection

Official ProLIP weights, Probabilistic Language-Image Pre-Training (ICLR 2025) • 7 items • Updated Apr 18, 2025 • 10

upvoted a paper almost 2 years ago

MaskRIS: Semantic Distortion-aware Data Augmentation for Referring Image Segmentation

Paper • 2411.19067 • Published Nov 28, 2024 • 8

upvoted a collection almost 2 years ago

Cosmos-Tokenizer1

Collection

⚠️ This collection is archived. 👉 https://huggingface.co/collections/nvidia/cosmos3 • 22 items • Updated Aug 11 • 44