Sergiu Han

hgsg's activity

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

upvoted a collection 1 day ago

MiMo-V2.6

Collection

3 items • Updated 1 day ago • 84

upvoted a paper 2 days ago

DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression

Paper • 2609.19969 • Published 6 days ago • 171

upvoted 10 papers 8 days ago

DSpark: Confidence-Scheduled Speculative Decoding with Semi-Autoregressive Generation

Paper • 2607.05147 • Published Jul 6 • 50

DeepSeek-V4: Towards Highly Efficient Million-Token Context Intelligence

Paper • 2606.19348 • Published Apr 26 • 45

Native Sparse Attention: Hardware-Aligned and Natively Trainable Sparse Attention

Paper • 2502.11089 • Published Feb 16, 2025 • 171

mHC: Manifold-Constrained Hyper-Connections

Paper • 2512.24880 • Published Dec 31, 2025 • 336

DeepSeek-OCR 2: Visual Causal Flow

Paper • 2601.20552 • Published Jan 28 • 75

Conditional Memory via Scalable Lookup: A New Axis of Sparsity for Large Language Models

Paper • 2601.07372 • Published Jan 12 • 53

DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models

Paper • 2402.03300 • Published Feb 5, 2024 • 150

DeepSeek-V3 Technical Report

Paper • 2412.19437 • Published Dec 27, 2024 • 89

DeepSeek-V2: A Strong, Economical, and Efficient Mixture-of-Experts Language Model

Paper • 2405.04434 • Published May 7, 2024 • 30

DeepSeekMoE: Towards Ultimate Expert Specialization in Mixture-of-Experts Language Models

Paper • 2401.06066 • Published Jan 11, 2024 • 65

upvoted a collection 16 days ago

MiniCPM5

Collection

SOTA on-device LLMs, small yet powerful. • 24 items • Updated 8 days ago • 64

upvoted a collection 3 months ago

DeepSpec

Collection

12 items • Updated Jun 28 • 94

upvoted 2 papers 3 months ago

LoopCoder-v2: Only Loop Once for Efficient Test-Time Computation Scaling

Paper • 2606.18023 • Published Jun 16 • 161

VisualClaw: A Real-Time, Personalized Agent for the Physical World

Paper • 2606.16295 • Published Jun 15 • 29

upvoted a paper 4 months ago

Harness Updating Is Not Harness Benefit: Disentangling Evolution Capabilities in Self-Evolving LLM Agents

Paper • 2605.30621 • Published May 28 • 22

upvoted a collection 4 months ago

Mistral Medium 3.5

Collection

Our first flaship models handling instruction-following, reasoning, and coding in a single set of opened-weights. • 2 items • Updated Apr 29 • 22

upvoted 2 papers 4 months ago

AutoMedBench: Towards Medical AutoResearch with Agentic AI Models

Paper • 2606.01961 • Published Jun 3 • 27

StepAudio 2.5 Technical Report

Paper • 2605.23463 • Published May 22 • 49