HackerTrans
TopNewTrendsCommentsPastAskShowJobs

tims457

129 karmajoined قبل 5 سنوات
https://tims457.github.io/ https://twitter.com/tims457

Submissions

Yi Ma – Pursuing the Nature of Intelligence

youtube.com
1 points·by tims457·قبل 4 أيام·0 comments

NeuroMechFly v2: simulating embodied sensorimotor control in adult Drosophila

nature.com
2 points·by tims457·قبل 4 أشهر·1 comments

Neural Algorithmic Reasoning

algo-reasoning.github.io
3 points·by tims457·السنة الماضية·0 comments

Ask HN: Alternative to Causal for probabilistic spreadsheet models

2 points·by tims457·قبل سنتين·1 comments

The Thinking Game

tribecafilm.com
1 points·by tims457·قبل سنتين·0 comments

[untitled]

1 points·by tims457·قبل سنتين·0 comments

Friends from the Old Neighborhood Turn Rivals in Big Tech's A.I. Race

nytimes.com
2 points·by tims457·قبل سنتين·0 comments

CS224W: Machine Learning with Graphs – Notes

snap-stanford.github.io
3 points·by tims457·قبل سنتين·0 comments

Graph Convolutional Networks

tkipf.github.io
3 points·by tims457·قبل سنتين·0 comments

Collective intelligence for deep learning: A survey of recent developments

journals.sagepub.com
2 points·by tims457·قبل سنتين·0 comments

Ask HN: What are the best resources you've found on improving your writing?

1 points·by tims457·قبل 3 سنوات·1 comments

comments

tims457
·قبل 4 أشهر·discuss
https://www.biorxiv.org/content/10.1101/2023.09.18.556649v4
tims457
·قبل 3 سنوات·discuss
No transparency here. What is it considering to come up with these scores? Are they prioritizing growth, value, financial health, or something else?
tims457
·قبل 3 سنوات·discuss
Haven't read the paper yet, but it's clear that the brain is much more sample efficient than artificial neural networks. We should be spending more time on figuring out why than scaling up LLMs.
tims457
·قبل 3 سنوات·discuss
I'm risking just posting another process that you might have seen, but I thought it was helpful. Maybe you or someone else will too.

Farnam Street annual review https://fs.blog/annual-review/ (Direct link without giving your email: https://fsmisc.s3.ca-central-1.amazonaws.com/2022+AR.pdf)