Eugene Oskin's picture

Eugene Oskin

eoskin

·

AI & ML interests

None yet

Recent Activity

liked a Space about 9 hours ago

AdithyaSK/rl-environments-guide

liked a model 2 days ago

jedisct1/MiMo-V2.5-coder-Q2

liked a Space 7 days ago

HuggingFaceBio/carbon-demo

View all activity

Organizations

None yet

upvoted 2 papers 11 days ago

SANA-WM: Efficient Minute-Scale World Modeling with Hybrid Linear Diffusion Transformer

Paper • 2605.15178 • Published 14 days ago • 83

Self-Distilled Agentic Reinforcement Learning

Paper • 2605.15155 • Published 14 days ago • 110

upvoted a paper 13 days ago

DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter

Paper • 1910.01108 • Published Oct 2, 2019 • 23

upvoted a collection 13 days ago

Qwen3.6

4 items • Updated Apr 22 • 380

upvoted 3 papers 14 days ago

MemPrivacy: Privacy-Preserving Personalized Memory Management for Edge-Cloud Agents

Paper • 2605.09530 • Published 18 days ago • 145

Efficient Pre-Training with Token Superposition

Paper • 2605.06546 • Published 21 days ago • 45

MoESD: Unveil Speculative Decoding's Potential for Accelerating Sparse MoE

Paper • 2505.19645 • Published Feb 16 • 1

upvoted an article 14 days ago

Article

What’s MXFP4? The 4-Bit Secret Powering OpenAI’s GPT‑OSS Models on Modest Hardware

RakshitAralimatti

•

Aug 8, 2025

• 35

upvoted 2 papers 15 days ago

SD-MoE: Spectral Decomposition for Effective Expert Specialization

Paper • 2602.12556 • Published Feb 13 • 1

Mean Mode Screaming: Mean--Variance Split Residuals for 1000-Layer Diffusion Transformers

Paper • 2605.06169 • Published 21 days ago • 228

upvoted an article 16 days ago

Article

KV Caching Explained: Optimizing Transformer Inference Efficiency

not-lain

•

Jan 30, 2025

• 340

upvoted a collection 20 days ago

🍎 Qwopus3.6

This collection features the advanced Qwopus3.6 series of multimodal large models, which are fine-tuned from the Qwen3.6 base models with a focus on e • 10 items • Updated 4 days ago • 55

upvoted a paper 21 days ago

Thinking with Drafting: Optical Decompression via Logical Reconstruction

Paper • 2602.11731 • Published Feb 12 • 36

upvoted a collection 23 days ago

Granite 4.1 Language Models

Efficient language models for multilingual generation, coding, RAG, and AI assistant workflows. • 6 items • Updated 28 days ago • 53

upvoted an article 23 days ago

Article

Granite 4.1 LLMs: How They’re Built

ibm-granite

•

28 days ago

• 75

upvoted 2 papers 24 days ago

Embedded Rust or C Firmware? Lessons from an Industrial Microcontroller Use Case with Ariel OS

Paper • 2604.25679 • Published 30 days ago • 1

Hierarchical Reasoning Model

Paper • 2506.21734 • Published Jun 26, 2025 • 54

upvoted a collection 26 days ago

Qwen-Scope

16 items • Updated 13 days ago • 68

upvoted a paper 28 days ago

Learning Transferable Visual Models From Natural Language Supervision

Paper • 2103.00020 • Published Feb 26, 2021 • 22

upvoted an article 30 days ago

Article

How 🤗 Accelerate runs very large models thanks to PyTorch

sgugger

•

Sep 27, 2022

• 18