Hacker News
new
|
past
|
comments
|
ask
|
show
|
jobs
|
submit
|
E-Reverance's submissions
login
1.
Recycler: Cost-Free Latent Reasoning via Hidden State Recycling [pdf]
(
bytednsdoc.com
)
1 point
by
E-Reverance
2 days ago
|
past
|
discuss
2.
Accelerating Attention for Ultra-Long Context
(
prefrontal.ai
)
3 points
by
E-Reverance
4 days ago
|
past
|
discuss
3.
Make Quizzes with Your Friends
(
instagram.com
)
1 point
by
E-Reverance
10 days ago
|
past
|
1 comment
4.
Diffeomorphic Optimization
(
arxiv.org
)
2 points
by
E-Reverance
16 days ago
|
past
5.
Parallel Language Models via Variational Decoding
(
arxiv.org
)
4 points
by
E-Reverance
17 days ago
|
past
6.
University of Leicester to make redundant all pure math professors (2021)
(
reddit.com
)
24 points
by
E-Reverance
20 days ago
|
past
|
2 comments
7.
Next Concept Models
(
arxiv.org
)
4 points
by
E-Reverance
22 days ago
|
past
8.
Correction Training for Diffusion Models
(
sunovivid.github.io
)
2 points
by
E-Reverance
24 days ago
|
past
9.
Leading Chinese mathematicians unite behind use of AI in research
(
scmp.com
)
4 points
by
E-Reverance
24 days ago
|
past
10.
The Unreasonable Effectiveness of Linear Prediction as a Perceptual Metric
(
arxiv.org
)
1 point
by
E-Reverance
26 days ago
|
past
11.
Residual Matrix Transformers: Scaling the Size of the Residual Stream
(
arxiv.org
)
3 points
by
E-Reverance
27 days ago
|
past
12.
Skate 3 in the Browser
(
aaddpp.lol
)
2 points
by
E-Reverance
30 days ago
|
past
|
2 comments
13.
Discrete Fourier Transform faster than n log n [pdf]
(
github.com/openai
)
2 points
by
E-Reverance
30 days ago
|
past
14.
Integer multiplication below n log n
(
github.com/openai
)
113 points
by
E-Reverance
30 days ago
|
past
|
83 comments
15.
Dust: Pretraining Transformers Without Backpropagation
(
qlabs.sh
)
280 points
by
E-Reverance
31 days ago
|
past
|
82 comments
16.
Single cortical neurons as deep artificial neural networks
(
sciencedirect.com
)
2 points
by
E-Reverance
33 days ago
|
past
17.
SoftServe: A Scalable Quasi-Newton Method for Deep Learning
(
arxiv.org
)
1 point
by
E-Reverance
34 days ago
|
past
18.
Chess, but it looks like doodles on paper
(
twitter.com/kevin_t_ngo
)
2 points
by
E-Reverance
36 days ago
|
past
19.
Self-Play Pretraining with Zero Data
(
arxiv.org
)
4 points
by
E-Reverance
42 days ago
|
past
20.
Learning to Discover Interesting Mathematics
(
arxiv.org
)
1 point
by
E-Reverance
42 days ago
|
past
21.
Log-Depth Recurrent Language Modeling
(
arxiv.org
)
1 point
by
E-Reverance
43 days ago
|
past
22.
Realtime Pathtracing in Octane
(
twitter.com/icelaglace
)
1 point
by
E-Reverance
44 days ago
|
past
23.
94% on AIME with 1B Params
(
paradigma.inc
)
2 points
by
E-Reverance
45 days ago
|
past
|
1 comment
24.
86" 3D Display
(
lookingglassfactory.com
)
3 points
by
E-Reverance
46 days ago
|
past
25.
Fluid Simulation in One Tweet
(
davidar.io
)
1 point
by
E-Reverance
46 days ago
|
past
26.
Solving rubiks cubes "without search"
(
arxiv.org
)
4 points
by
E-Reverance
48 days ago
|
past
27.
Jev might be Qwen finetune
(
nitter.click
)
2 points
by
E-Reverance
49 days ago
|
past
28.
DeepSeek's Parent Company
(
high-flyer.cn
)
1 point
by
E-Reverance
50 days ago
|
past
29.
Replicating Jev with Qwen-2.5
(
perennialte.ch
)
3 points
by
E-Reverance
51 days ago
|
past
30.
Dart: Denoising Autoregressive Transformer
(
arxiv.org
)
1 point
by
E-Reverance
52 days ago
|
past
More
Guidelines
|
FAQ
|
Lists
|
API
|
Security
|
Legal
|
Apply to YC
|
Contact
Search: