Articles by dataminer
95

A 30B Qwen model walks into a Raspberry Pi and runs in real time (byteshape.com)

2

Court Filings Allege Meta Downplayed Risks to Children and Misled the Public (time.com)

2

Shattering the Illusion: Maker Achieves Million-Step, Zero-Error LLM Reasoning (cognizant.com)

46

HipKittens: Fast and furious AMD kernels (stanford.edu)

2

Pretraining with hierarchical memories separating long-tail and common knowledge (arxiv.org)

2

China's largest chipmaker testing first homegrown immersion DUV litho tool (tomshardware.com)

2

Canada became the centre of a measles outbreak in North America (bbc.com)

1

Memoir: Lifelong Model Editing with Minimal Overwrite Informed Retention for LLM (arxiv.org)

1

The Earliest Language Model (colab.research.google.com)

7

Ai2 Paper Finder (allen.ai)

1

Transformers Without Normalization (jiachenzhu.github.io)

4

GeForce RTX 5090 fails to topple RTX 4090 in GPU compute benchmark (notebookcheck.net)

1

FireDucks Compiler Accelerated DataFrame Library for Python (fireducks-dev.github.io)

4

DeepSeek-R1, Q8 Quant (Max Quality). 768GB of DDR5, 2x AMD EPYC CPU, No GPU (twitter.com/carrigmat)

3

Huawei has a new chip to rival Nvidia's AI GPUs – Ascend 910C (tomshardware.com)

2

What the First All-EV Car Market Looks Like (bloomberg.com)

169

Scientist treated her own cancer with viruses she grew in the lab (nature.com)

105

Zamba2-7B (zyphra.com)

1

Mike Little (wikipedia.org)

1

Day Trader Says He Made $306M on Tesla, Then Lost It All (bloomberg.com)

71

OpenAI Discusses Giving Altman 7% Stake in For-Profit Shift (bloomberg.com)

2

Macklem Warns AI May Push Prices Higher Through Demand Boost (bnnbloomberg.ca)

38

DifuzCam: Replacing Camera Lens with a Mask and a Diffusion Model (arxiv.org)

1

Robots and Specialized Software Open New Solar Farm Development Opportunities (bloomberg.com)

1

The Geometry of Categorical and Hierarchical Concepts in Large Language Models (arxiv.org)

2

Python 3.12 Preview: Subinterpreters (realpython.com)

2

Have live audio/video conversations with ChatGPT (livekit.io)

3

Adding NVMe SSDs to Enable and Accelerate 100B Model Fine-Tuning on a Single GPU (arxiv.org)

2

Humanoid Locomotion as Next Token Prediction (humanoid-next-token-prediction.github.i...

3

How do Spotify Codes work? (boonepeter.github.io)

1

Variational autoencoder for design of synthetic viral vector serotypes (nature.com)

1

How AI is decoding the animal kingdom (ft.com)

0

Unified diffs make GPT-4 Turbo less lazy (aider.chat)

2

LLM Augmented LLMs: Expanding Capabilities Through Composition (arxiv.org)

2

2084: MarcRandbot: Speech Synthesis with Mamba (2084.substack.com)

2

Losing weight without any stress (smartbmicalculator.com)

3

Degrowth can work here’s how science can help (nature.com)

99

High-Speed Large Language Model Serving on PCs with Consumer-Grade GPUs (github.com/sjtu-ipads)

1

Gated Linear Attention Transformers with Hardware-Efficient Training (arxiv.org)

32

Every Bitcoin payment 'uses a swimming pool of water' (bbc.com)

3

Trends in Deep Learning Hardware: Bill Dally (Nvidia) [video] (youtube.com)

1

Self-Refine: Iterative Refinement with Self-Feedback (selfrefine.info)

1

Machine Learning Engineering Guides and Tools (github.com/stas00)

4

In London, the Long Wait for a Super Sewer Is Almost Over (bloomberg.com)

20

Auto-Regressive Next-Token Predictors Are Universal Learners (arxiv.org)

2

Parameter-Free Molecular Classification and Regression with Gzip (chemrxiv.org)

3

Eli Lilly experimental obesity drug helped patients lose upto 24% of weight (cnbc.com)

1

The testing framework dedicated to ML models, from tabular to LLMs (github.com/giskard-ai)

1

The Rate of Return on Everything, 1870–2015 [pdf] (economics.harvard.edu)

1

Hybrid Electric Powered Multi-Lobed Airship for Sustainable Aviation (mdpi.com)

7

Ellison's Cave (wikipedia.org)

3

Virus-Like “Jumping” DNA Regulates Human Neurons (scitechdaily.com)

181

[flagged] The 1% of the population accountable for 63 % of all violent crime convictions (2013) (nih.gov)

115

The Little Book of Deep Learning [pdf] (fleuret.org)

2

Recurrent Memory Transformer [pdf] (neurips.cc)

1

AI is already taking video game illustrators’ jobs in China (restofworld.org)

75

[flagged] Abortion Pill Blocked Nationwide Under Order from US Judge (bloomberg.com)

3

Extending Communications – Drone-Based Radio Repeater (hackster.io)

2

Italo Calvino ' the Black Sheep ' (aladin.co.kr)

3

The race to make diesel engines run on hydrogen (bbc.com)

3

A secret message in a Colombian song gave hostages hope (bbc.com)

118

Pomsky – A portable, modern regular expression language (pomsky-lang.org)

4

Russia burning huge amounts of gas, puzzling experts

2

SARS-CoV-2:spike protein mutational analysis, epitope for broad neutralization

2

France tweaks rules to keep nuclear plants running during heatwave