Mathias Nielsen

mathiasn1

MediaCatch

·

https://grandaiwizard.com/

AI & ML interests

🏢 Senior Machine Learning Engineer @ https://mediacatch.io/

Recent Activity

updated a dataset 1 day ago

MediaCatch/coral-v3

updated a dataset 1 day ago

MediaCatch/mc_da_spontaneous

updated a dataset 1 day ago

MediaCatch/nota

View all activity

Organizations

upvoted a paper 4 days ago

Granary: Speech Recognition and Translation Dataset in 25 European Languages

Paper • 2505.13404 • Published May 19, 2025 • 4

upvoted a collection 23 days ago

Muse Spark Open

Muse Spark Open model. A 816B parameter model with 42B Active parameters. BF16 and NVFP4 Weights. • 2 items • Updated 23 days ago • 176

upvoted a paper 24 days ago

UltraBERT: Beyond the Limit of BERT Pre-Training

Paper • 2610.15028 • Published 29 days ago • 63

upvoted a collection 24 days ago

UltraBERT

UltraBERT: Beyond the Limit of BERT Pre-Training. Pre-trained 150M/380M encoders and a text embedding model, English, 32K context. • 4 items • Updated 25 days ago • 32

upvoted an article about 2 months ago

Article

Know Who Spoke When: Build Real-Time, Multi-Speaker AI with NVIDIA Nemotron 3 Diarization

nvidia

•

Sep 23

• 73

upvoted 2 collections about 2 months ago

Qwen3.8

4 items • Updated Aug 13 • 611

Qwen3.6

4 items • Updated Apr 22 • 507

upvoted a paper 3 months ago

DFM Mimir v1: An Open HRM Delivering Frontier Performance at 1B Parameters Using Only Permissible Post-Training Data

Paper • 2608.13517 • Published Aug 13 • 32

upvoted 2 articles 3 months ago

Article

mDenseOn with the mLateOn: Open Multilingual, Long-Context, and Code Retrieval Models

lightonai

•

Jul 30

• 41

Article

GPU Management: Why Idle GPUs Are the New Grounded Aircraft

Dharma-AI

•

Jul 30

• 97

upvoted 3 articles 4 months ago

Article

Security incident disclosure — July 2026

system

•

Jul 16

• 814

Article

Model Routing Is Simple. Until It Isn’t.

ibm-research

•

Jul 15

• 95

Article

What building Shippy taught us about building agents

allenai

•

Jul 15

• 23

upvoted a collection 4 months ago

Gemma 4

Gemma 4 is Google's new model family including including E2B, E4B, 26B-A4B, and 31B. • 43 items • Updated 11 days ago • 274

upvoted 2 articles 4 months ago

Article

Welcome Inkling by Thinking Machines

+3

burtenshaw, merve, pcuenq, ariG23498, andito

•

Jul 15

• 170

Article

How to Fine-Tune Nemotron 3.5 ASR for Your Language, Domain, or Accent

nvidia

•

Jun 4

• 82

upvoted 2 articles 5 months ago

Article

Introducing BERTopic Integration with the Hugging Face Hub

MaartenGr, davanstrien

•

May 31, 2023

• 12

Article

We got local models to triage the OpenClaw repo for FREE!*

+1

osolmaz, burtenshaw, evalstate

•

Jun 22

• 26

upvoted a paper 5 months ago

PP-OCRv6: From 1.5M to 34.5M Parameters, Surpassing Billion-Scale VLMs on OCR Tasks

Paper • 2606.13108 • Published Jun 11 • 10

upvoted an article 5 months ago

Article

PP-OCRv6 on Hugging Face: 50-Language OCR from 1.5M to 34.5M Parameters

PaddlePaddle

•

Jun 22

• 41