E-Reverance

9 ms·

91.
Meshtryoshka: Differentiable Mesh Rendering for Unbounded Scenes(danielxu9393.github.io)
2 points by E-Reverance 3mo ago|0 comments
92.
Simplified Sparse Attention via Gist Tokens(arxiv.org)
4 points by E-Reverance 3mo ago|0 comments
93.
Continuity-Enhancing Degree Elevation and Splits(graphics.cs.utah.edu)
2 points by E-Reverance 3mo ago|0 comments
94.
Forward Self Models(jagilley.github.io)
2 points by E-Reverance 3mo ago|0 comments
95.
Tapered Language Models(arxiv.org)
3 points by E-Reverance 3mo ago|0 comments
96.
Everything is logarithms(alexkritchevsky.com)
316 points by E-Reverance 3mo ago|83 comments
97.
by E-Reverance 3mo ago
I don't think the issue is that they used ai behind the scenes, but there is an implicit proof of work from forcing it beyond the style you'd expect. I for one roll my eyes whenever I see that specific kind of rounded corner, fros
98.
Explaining Attention with Program Synthesis(arxiv.org)
2 points by E-Reverance 3mo ago|0 comments
99.
by E-Reverance 3mo ago
"At the same time, China is also the world's leading producer of electric cars..." Kind of interesting for a professionally branded company to use "..." like that
100.
by E-Reverance 3mo ago
I think its worth emphasizing that his argument isn't completely against generative ai, but rather its environment. Although I don't see why it would be impossible for something like an LLM to learn some sort of self-play within i
101.
by E-Reverance 3mo ago
I don't completely disagree but its worth noting how new a lot of the empirical evidence in favour of LLMs are, so its not impossible to be a tad ignorant of the present
102.
by E-Reverance 3mo ago
No > It uses 384 routed experts (top-8) with hybrid attention (full-attention + sliding-window 128 at 6:1 ratio) over 70 layers (1 dense + 69 MoE) https://recipes.vllm.ai/XiaomiMiMo/MiMo-V2.5-Pro
103.
Sparsely gated tiny linear experts(arxiv.org)
3 points by E-Reverance 3mo ago|0 comments
104.
Training-Free Single-Image Diffusion Models(haojunqiu.github.io)
6 points by E-Reverance 4mo ago|0 comments
105.
Simple input–output dependencies explain neuronal activity(nature.com)
4 points by E-Reverance 4mo ago|0 comments
106.
Self Teaching Autoencoder(the-puzzler.github.io)
2 points by E-Reverance 4mo ago|0 comments
107.
by E-Reverance 4mo ago
> P.S.: I like discussing such topics. If anyone knows a forum or discord with like-minded people, please let me know :) Unironically twitter (and only use the "Following" tab as opposed to the "For You") Make an acco
108.
by E-Reverance 4mo ago
> But I am a bit reasurred that at least my job won't be fully replaced with AI :) I honestly can't comment with certainty that training from videos alone and whatever tokenization scheme they're using will ever get perfec
109.
Distance Marching for Generative Modeling(arxiv.org)
1 points by E-Reverance 4mo ago|0 comments
110.
A Theory of Generalization in Deep Learning(arxiv.org)
4 points by E-Reverance 5mo ago|0 comments
111.
by E-Reverance 5mo ago
FYI it’s not an Arab country
112.
by E-Reverance 5mo ago
Interesting comment from him: " SPIEGEL: So you don't consider Collins to be a true scientist? Venter: Let's just say he's a government administrator. " https://www.science.org/content/blog-post
113.
Neural similarity predicts whether strangers become friends(nature.com)
4 points by E-Reverance 5mo ago|0 comments
114.
Generation Is Required for Data-Efficient Perception(arxiv.org)
1 points by E-Reverance 5mo ago|0 comments
115.
by E-Reverance 5mo ago
They factorize the distribution in which they are trained on which is essentially generalization https://arxiv.org/abs/2602.02385
116.
by E-Reverance 5mo ago
Yeah I wouldn't be surprised if journalist are getting high on their own supply of resentment and fear mongering
117.
by E-Reverance 5mo ago
People not wanting their jobs be automated is different from not yearning for automation as a principle . Most people want or (at least don't mind) elevators, tap water, dishwashers, traffic lights, electrical fuses, sliding doors
118.
by E-Reverance 5mo ago
It does? " Software brain is powerful stuff. It’s a way of thinking that basically created our modern world. Marc Andreessen, the literal embodiment of software brain, called it in 2011 when he wrote the piece “Why software is eating t
119.
by E-Reverance 5mo ago
Maybe a nitpicky HN comment, but why are we lumping the term automation with very recent grievances about certain kinds of automation
120.
by E-Reverance 5mo ago
Try refreshing or something, still works for me

More ›