Cosmo's picture

Cosmo

cosmojg

·

https://cosmo.red

AI & ML interests

Machine learning and computational neuroscience

Recent Activity

liked a model about 3 hours ago

Qwen/Qwen3-4B-Instruct-2507

liked a model about 3 hours ago

Qwen/Qwen3-4B-Thinking-2507

liked a Space about 3 hours ago

Qwen/Qwen3-Coder-WebDev

View all activity

Organizations

None yet

upvoted a collection 2 days ago

GLM-4.5

10 items • Updated 3 days ago • 191

upvoted a paper 24 days ago

Robust Learning of Diverse Code Edits

Paper • 2503.03656 • Published Mar 5 • 3

upvoted 4 collections 24 days ago

NextCoder

NextCoder family of code-editing LMs developed with Selective Knowledge Transfer and its training data. • 6 items • Updated 28 days ago • 69

T5Gemma

32 items • Updated 28 days ago • 60

MedGemma Release

Collection of Gemma 3 variants for performance on medical text and image comprehension to accelerate building healthcare-based AI applications. • 7 items • Updated 26 days ago • 271

Kimi-K2

Moonshot's MoE LLMs with 1 trillion parameters, exceptional on agentic intellegence • 2 items • Updated 25 days ago • 115

upvoted an article 24 days ago

Article

Reachy Mini - The Open-Source Robot for Today's and Tomorrow's AI Builders

By

and 1 other •

29 days ago

• 636

upvoted an article 29 days ago

Article

SmolLM3: smol, multilingual, long-context reasoner

By

and 22 others •

30 days ago

• 611

upvoted a paper about 1 month ago

Pangu Ultra MoE: How to Train Your Big MoE on Ascend NPUs

Paper • 2505.04519 • Published May 7 • 4

upvoted a collection about 1 month ago

Jamba 1.7

The AI21 Jamba family of models are hybrid SSM-Transformer foundation models, blending speed, efficient long context processing, and accuracy. • 4 items • Updated Jul 2 • 10

upvoted a paper about 1 month ago

MiniMax-M1: Scaling Test-Time Compute Efficiently with Lightning Attention

Paper • 2506.13585 • Published Jun 16 • 260

upvoted a collection about 1 month ago

MiniMax-M1

MiniMax-M1, the world's first open-weight, large-scale hybrid-attention reasoning model. • 6 items • Updated Jul 3 • 110

upvoted a paper about 1 month ago

SynLogic: Synthesizing Verifiable Reasoning Data at Scale for Learning Logical Reasoning and Beyond

Paper • 2505.19641 • Published May 26 • 67

upvoted a collection about 1 month ago

SynLogic

Synthesizing Verifiable Reasoning Data at Scale for Learning Logical Reasoning and Beyond • 5 items • Updated Jun 3 • 11

upvoted a collection about 2 months ago

V-JEPA 2

A frontier video understanding model developed by FAIR, Meta, which extends the pretraining objectives of https://ai.meta.com/blog/v-jepa-yann • 8 items • Updated Jun 13 • 153

upvoted 3 collections 2 months ago

Helium 1

Helium 1: a modular and multilingual LLM • 9 items • Updated Apr 30 • 9

Moshi v0.1 Release

MLX, Candle & PyTorch model checkpoints released as part of the Moshi release from Kyutai. Run inference via: https://github.com/kyutai-labs/moshi • 15 items • Updated Apr 18 • 235

Hibiki fr-en

Hibiki is a model for streaming speech translation , which can run on device! See https://github.com/kyutai-labs/hibiki. • 5 items • Updated Feb 6 • 53

upvoted an article 3 months ago

Article

The Open Medical-LLM Leaderboard: Benchmarking Large Language Models in Healthcare

By

and 2 others •

Apr 19, 2024

• 174

upvoted a collection 3 months ago

Phi-4

Phi-4 family of small language, multi-modal and reasoning models. • 17 items • Updated 27 days ago • 176