HackerTrans
トップ新着トレンドコメント過去質問紹介求人

seccode

no profile record

投稿

Tell HN: Perfect – not statistical – ZKP for Kolmogorov complexity

1 ポイント·投稿者 seccode·昨年·2 コメント

Compressor "firn" improves ZSTD by 13%

github.com
3 ポイント·投稿者 seccode·2 年前·3 コメント

[untitled]

7 ポイント·投稿者 seccode·2 年前·0 コメント

The digits of pi are not random

github.com
10 ポイント·投稿者 seccode·2 年前·24 コメント

Show HN: New compression algorithm beats ZSTD by 14%

github.com
1 ポイント·投稿者 seccode·2 年前·3 コメント

Show HN: KeyPush: Mix of Bing/ChatGPT

keypush.ai
1 ポイント·投稿者 seccode·2 年前·1 コメント

Full Self-Driving Struggles

bloomberg.com
3 ポイント·投稿者 seccode·2 年前·0 コメント

Show HN: Git Searcher, search commits for an added or removed string

github.com
3 ポイント·投稿者 seccode·2 年前·8 コメント

[untitled]

3 ポイント·投稿者 seccode·2 年前·0 コメント

コメント

seccode
·昨年·議論
There is a well-known paper related to a statistical zero-knowledge proof about Kolmogorov complexity, but this proof introduced is considered a perfect ZKP
seccode
·昨年·議論
I think I heard that 77 was factored as well
seccode
·2 年前·議論
Maybe truth here, but also Microsoft didn't lead their latest round, which isn't a great sign for their moat
seccode
·2 年前·議論
Contract law research
seccode
·2 年前·議論
Methodology for comparison: train zstd dictionary on enwik9. Then build my dictionary as most common words in enwik9. Mine does 13% better because of the way I discovered how you can generate dictionary replacement symbols
seccode
·2 年前·議論
This is a direct comparison to zstd with a dictionary actually
seccode
·2 年前·議論
Syracuse airport. You're telling me just because there's 2 feet of snow on the runway that you can't fly?
seccode
·2 年前·議論
Getting higher z-scores now. But you could always just try running the code yourself
seccode
·2 年前·議論
I used the z-score. How can you claim that the digits of pi are random, yet a random forest classifier predicted better than the distribution probability. Your claim implicitly means "there is no structure." The hard thing to understand is that the classifier didn't see the test set, so what structure did it learn? At the very least this is an interesting question
seccode
·2 年前·議論
Better to use statistical significance tests to talk about what is "far more likely"
seccode
·2 年前·議論
It doesn't predict better than even, it predicts better than the distribution probability
seccode
·2 年前·議論
This doesn't _break_ sha256, just opens the door to breaking sha256 with machine learning
seccode
·2 年前·議論
Old people not listening to young people
seccode
·2 年前·議論
Would greatly appreciate any peer review for this work
seccode
·2 年前·議論
My dad didn't use it for a long time but has used it now and then recently and has found it to be very impressive
seccode
·2 年前·議論
Thanks this is a good point, I'll change proof to evidence
seccode
·2 年前·議論
I'm not predicting the number I'm predicting number%2==0. The model predicted better than the distribution probability
seccode
·2 年前·議論
Thanks for teaching me some important statistics!
seccode
·2 年前·議論
Also, I am testing different ranges of digits other than first 10,000, but the problem with other ranges is that the distribution of digits is highly imbalanced and the model is not showing statistical significance, but models have a harder time when the distribution of classes is not 50/50, so I think its not quite fair to evaluate the model on these ranges.

So why do you think the first 10,000 digits are somewhat predictable?
seccode
·2 年前·議論
The issue is not with getting the digits, the issue is with running a large model for larger digit ranges. I tried running with 10,000,000 digits and haven't gotten a prediction yet.