HackerTrans
TopNewTrendsCommentsPastAskShowJobs

moondistance

no profile record

Submissions

VibeThinker-3B achieves 80.2 on LCBv6

twitter.com
8 points·by moondistance·il y a 28 jours·3 comments

Qwen 3.5: small models with impressive performance

twitter.com
6 points·by moondistance·il y a 4 mois·0 comments

Language Models Are Injective and Hence Invertible

arxiv.org
1 points·by moondistance·il y a 9 mois·2 comments

Autonomous Trash Can

twitter.com
20 points·by moondistance·il y a 12 mois·1 comments

Grok 4 Benchmarks

twitter.com
3 points·by moondistance·l’année dernière·0 comments

Jensen Huang – Nvidia GTC 2025 Keynote

nvidia.com
75 points·by moondistance·l’année dernière·74 comments

Exaone Deep 32B – beats DeepSeek 671B

lgresearch.ai
3 points·by moondistance·l’année dernière·1 comments

DeepSeek-R1 at 3,872 tokens / second on a single Nvidia HGX H200

blogs.nvidia.com
13 points·by moondistance·l’année dernière·1 comments

ByteDance Doubao-1.5-pro matches GPT 4o benchmarks at 50x cheaper

twitter.com
4 points·by moondistance·l’année dernière·0 comments

US-China Commission top recommendation: Manhattan project for race to AGI [pdf]

uscc.gov
4 points·by moondistance·il y a 2 ans·0 comments

AI scans RNA 'dark matter' and uncovers 70k new viruses

nature.com
1 points·by moondistance·il y a 2 ans·0 comments

comments

moondistance
·il y a 28 jours·discuss
Paper: https://huggingface.co/papers/2606.16140
moondistance
·il y a 11 mois·discuss
How does this compare with Exowatt? Assume fresnel lenses are more efficient.
moondistance
·l’année dernière·discuss
Chris Crawford is also famous for the Dragon Speech :) https://youtu.be/CBrj4S24074?si=Ph12RpW8BKsh8-qS
moondistance
·l’année dernière·discuss
Yes, big time, but there continues to be lots of progress.

Most importantly, models are maturing, and this means less custom optimization is required.
moondistance
·l’année dernière·discuss
(this is also why AMD popped 10% at open yesterday - this is a new development and talks from their 2025 "Advancing AI" event were published late last week + over the weekend)
moondistance
·l’année dernière·discuss
Yes, for many applications.

Meta, OpenAI, Crusoe, and xAI recently announced large purchases of MI300 chips for inference.

MI400, which will be available next year, also looks to be at least on par with Nvidia's roadmap.
moondistance
·l’année dernière·discuss
This is from 2020.
moondistance
·l’année dernière·discuss
Thanks :)
moondistance
·l’année dernière·discuss
Where do you find/buy land? How do you vet purchases? Can you point to a few websites, etc.? Thanks!
moondistance
·il y a 2 ans·discuss
Happy New Year!

I know so few of you but feel like you’re my best friends. Love to all.
moondistance
·il y a 2 ans·discuss
Microsoft is negotiating equity in OpenAI as part of the switch to a for-profit. Non-zero chance this is a negotiation flex.