HK
Heykuki News
Top
New
Best
Ask
Show
Jobs
Toggle theme
Top
New
Best
Ask
Show
Jobs
Request
601.
▲
ChatGPT outperforms crowd-workers for text-annotation tasks
arxiv.org
205 comments
3 years ago
georgehill
240 points
602.
▲
Factoring 2048 RSA integers in 177 days with 13436 qubits and a multimode memory
arxiv.org
144 comments
5 years ago
athul7744
240 points
603.
▲
δ-mem: Efficient Online Memory for Large Language Models
arxiv.org
60 comments
a month ago
44za12
240 points
604.
▲
Neural Ordinary Differential Equations
arxiv.org
60 comments
8 years ago
asparagui
240 points
605.
▲
Learn Physics by Programming in Haskell [pdf]
arxiv.org
55 comments
11 years ago
lelf
240 points
606.
▲
The Machine That Builds Itself: The Strengths of the Lisp Languages (2016)
arxiv.org
318 comments
8 years ago
kuwze
239 points
607.
▲
Techno-feudalism and the rise of AGI: A future without economic rights?
arxiv.org
244 comments
a year ago
lexandstuff
239 points
608.
▲
LLMs cannot find reasoning errors, but can correct them
arxiv.org
142 comments
3 years ago
koie
239 points
609.
▲
Grokked Transformers Are Implicit Reasoners
arxiv.org
61 comments
2 years ago
jasondavies
239 points
610.
▲
The Principles of Diffusion Models
arxiv.org
31 comments
8 months ago
Anon84
239 points
611.
▲
Modern iOS Security Features – A Deep Dive into SPTM, TXM, and Exclaves
arxiv.org
32 comments
8 months ago
todsacerdoti
237 points
612.
▲
Ultra Fast Bert
arxiv.org
1 comment
3 years ago
gyre007
237 points
613.
▲
4.5M Suspected Fake Stars in GitHub
arxiv.org
212 comments
a year ago
qianli_cs
236 points
614.
▲
Google’s Neural Machine Translation System
arxiv.org
97 comments
10 years ago
fitzwatermellow
236 points
615.
▲
LoRA vs. Full Fine-Tuning: An Illusion of Equivalence
arxiv.org
53 comments
2 years ago
timbilt
236 points
616.
▲
How AI impacts skill formation
arxiv.org
5 comments
5 months ago
northfield27
236 points
617.
▲
Multiplying Matrices Without Multiplying
arxiv.org
122 comments
5 years ago
moinnadeem
235 points
618.
▲
Knowledge from small number of debates outperforms wisdom of large crowds (2017)
arxiv.org
116 comments
8 years ago
Dowwie
235 points
619.
▲
Regularized Newton Method with Global $O(1/k^2)$ Convergence
arxiv.org
103 comments
5 years ago
ColinWright
235 points
620.
▲
Scaling Transformers to 1B Tokens
arxiv.org
68 comments
3 years ago
mottiden
234 points
621.
▲
Thermodynamic Linear Algebra
arxiv.org
55 comments
3 years ago
aifer4
234 points
622.
▲
YOLOv4: Optimal Speed and Accuracy of Object Detection
arxiv.org
54 comments
6 years ago
groar
234 points
623.
▲
TinyLoRA – Learning to Reason in 13 Parameters
arxiv.org
45 comments
3 months ago
sorenjan
234 points
624.
▲
Matrix multiplication using only addition
arxiv.org
108 comments
3 years ago
daniel-cussen
233 points
625.
▲
Self-Compressing Neural Networks
arxiv.org
57 comments
2 years ago
bilsbie
233 points
626.
▲
New attention mechanisms that outperform standard multi-head attention
arxiv.org
49 comments
2 years ago
snats
233 points
627.
▲
Evaluating AGENTS.md: are they helpful for coding agents?
arxiv.org
161 comments
4 months ago
mustaphah
232 points
628.
▲
Catala: A Programming Language for the Law
arxiv.org
126 comments
5 years ago
todsacerdoti
232 points
629.
▲
Language models are injective and hence invertible
arxiv.org
148 comments
8 months ago
mazsa
231 points
630.
▲
XLSTMTime: Long-Term Time Series Forecasting with xLSTM
arxiv.org
53 comments
2 years ago
beefman
231 points
More