slybot

The language is not the problem here, I also cannot understand ancient greek. Let aside the casting, it impose a modern morality instead of the original moral arguments. The movie portrays Odysseus in a completely different manner. Homer's Odysseus is intelligent, deceitful, resilient, cruel... While in the movie Nolan's "modern" themes served rather than Homer's original.

I don't know if I am the only one, but it feels like trying to rewrite a historical epic with someone's own themes. I don't say one is better or worse, but it just doesn't feel right.

In this context, the book itself is the history rather than the story as a foundational western narrative. The movie diverge in many grounds from the book, and it is sad that most people will only watch the movie rather than read the original work.

Advertised as "Private way to connect" by new head, then WhatsApp is new Telegram?

Related: https://news.ycombinator.com/item?id=48632224 and https://x.com/kunalb11/status/2071624878713971176

Nive work! I solved all five puzzles and only annoyance was the quite a few singular+plural forms in the same puzzle. I expected all unique words.

Yes, did you? Anything Cypher, punk or cypherpunk inside that you can point out?

They do call themselves, "Partiya Karkerên Kurdistanê" (PKK) which literally means Kurdish Workers Party. It doesn't mean whether they call themselves terrorist or not; USA, UK, EU, NATO and many others call that way.

https://en.wikipedia.org/wiki/Kurdistan_Workers%27_Party

This is a work written by the terrorist organization leader after incarcerated on 1999. While this carefully selected piece doesn't, the larger work of five volumes named "Manifesto of the Democratic Civilization" goes into the manifestation of his terrorist movement.

Funnily, this small library features works outside of it's domain, including a manifesto from PKK terrorist organization leader..

This comment has a tone of external reviewer, while appearantly your profile shows you are affiliated.

Earlier discussion on this;

2024 - https://news.ycombinator.com/item?id=41047110 (41 comments)

2022 - https://news.ycombinator.com/item?id=33222687 (59 comments)

Well then same goes for saying, there was no genocide.

https://youtu.be/Rr9zXuG0-c0?si=O14GnPdhFXWKeMUm

Cannot agree more... Tired of flags being waved at the wrong places and times.

I did AoC 2021 until D10 using awk, it was fun but not easy and couldn't proceed further: https://github.com/nusretipek/Advent-of-Code-2021

https://m.imdb.com/title/tt3674140/

I watched this movie on cinema a decade ago. Highly recommended.

https://www.pcg-random.org/posts/bounded-rands.html

Couple of years back, I made a mistake on trusting Deutsche Bahn. I took a regional train from Düsseldorf, hoping to get at some connection hub (never reached and don't remember, possibly Cologne), I have to ICE to reach Brussels. It was evening and at some point the train stopped in the middle of nowhere and an announcement followed 20 minutes later, saying the train went through the wrong tracks and it cannot return! And, we need to wait couple of hours before the rails are cleared and then we will need to get of the next station, a tiny town where only few trains run per day and none after 10pm!. Genius idea to leave people there, conductor also puts his unsatisfaction because apparently he also needs to get off there and request a ride. Let alone, I missed my connections. I had to either wait there until morning or get a ride. Never used Deutsche Bahn again, I feel less stressed by driving, even I really enjoy train rides in general.

Did I get compensation? Yes I did after four months for the ticket price only (around 40 euros), but after ridiculous process of that I need to send them forms and tickets via regular post, with a stamp that can only post within Germany. Nice try by their side..

Funny how the article has no reference to C++ but still linking "Falling Into The Pit of Success". Did we all get our daily dose of subliminal messages from rust?

Recently, I also switched to Jellyfin. I still have access to Netflix and Disney through family plan. The service problem was the quality issue. I have the Ultra plan's while Netflix keep pushing SD (due to Widevine certificates). Simply cannot stand watching 480p on WQHD+ screen. For the content I have legitimate access, but cannot get good service, I don't consider it pirating.

->Bradley-Terry and Elo scores are equivalent mathematical models! No, they are not equivalent mathematical models, they are equalivant in terms of calculation of score function(logistic) given equivalent scale factors. Such that, Bradley-terry: 1/(1 + e^(x(r_B - r_A))) and Elo rating: 1/(1 + 10^((r_B - r_A)/y)), then equivalance requires x = ln(10)/y. More importantly, Elo rating is online scoring system, meaning it takes into accoun the sequence of the events. From your blog post, I understand that you are not updating the scores after after each event. In other words, Elo rating can be interpreted as an incremental fitting of a Bradley-Terry (using similar logistic) model but not the same!

-> The fundamental presumption is the same Thurstone model The Thurstone model is similar, and as you said it assumes normal (as opposed to logistic) using probit link function. It predates both models and due to computational constraints, you can call Bradley-Terry and Elo rating computationally convenient approximation of the Thurstone model.

-> We did experiment with a Bradley-Terry loss function (https://hackmd.io/eOwlF7O_Q1K4hj7WZcYFiw) The math is correct. Thanks for sharing. Indeed, if you do it with incremental updating, you will lose the differentiability given the next winning probability is dependent on the previous updates. Call it what you want, but note that this is not truly and Elo rating which leads misunderstanding. It is Bradley-Terry given you do batch updates which you take extra steps to connect with Elo score, as shown in the link.

Lastly, normal and logistic distribution will lead to log(0) in evaluations which results inf in loss. As I can see from you upper comment, you try add uniform(0.02) as ad-hoc fix. An elegant fix to that is use heavy-tailed distribution such as Cauchy.

They don't use Elo scores. See my comment above, the loss function is adopted from Bradley-Terry.

More confused,

1) 0.02 * random.random() != N(0, 0.02)

2) The LLM will sample a normal distribution, this only depends on your c parameter, the absolute scale doesn't matter neither in Bradley-Terry nor in Elo. So saying +-4 and claiming LLM reasoning in Standard normal is ridiculous.

3) > then we get a pairwise matrix with virtually identical statistical properties to the observed pairwise matrices. >>> then did you asked yourselves if I have "statistically identical" pair-wise matrix and observed pairwise matrix, the. why you even bother myself? You can simply use observed pairwise matrix...

Fit an ELO-style rating system (Bradley-Terry) to turn pairwise comparisons into absolute per-document scores.

There are some conceptual gaps and this sentence is misleading in general. First, this sentence implies that Bradley-Terry is a some sort of an Elo variant, which is not true. Elo rating introduced nearly 10 years later in completely different domain.

They are two completely different ranking systems. Bradley-Terry use ratio-based, while Elo use logistic score function. Scales of the scores are completely different as well as the their sensitivity to the score differences.

Possibly, Bradley-Terry is preffered by the authors due to simpler likelihood evaluation and update doesn't depend on the order of pairwise evaluations.

There is also variants of Elo-rating that use MLE (optimized Elo) and even recently Bayesian Elo. For post-hoc time invariant scores, there is randomized Elo rating and so on.

People like Elo ratings because they are simple to understand. Most of the time, they forget why they developed specifically for chess tournaments. All variants above and dozens more try to improve (fix) one aspect of the Elo ratings, because their application has no 100% clear determination of winner, the update scale parameter is too small or large, matches are played simultaneously, different matches played and so on.

Also, let say one document is always preffered one all LLMs then it has only wins, then MLE will result in flat marginal likelihood for that where the update parameter (c) will inf.

came here to say this, it is a ridiculuous number given ~70M deaf people in the world.

I always say the Turkish Lira follows the Argentina pattern with lag -3.5 years. https://www.investing.com/currencies/usd-ars

It is all about bad governance and leadership... Look at Argentina with half population and 3x more cultivatable. Don't surprise to see the parity >= 80 TRY / 1 USD in 2 years!

I have experienced similar discrimination while ago. It clearly shows some streotypes against certain nationalities. If it is not a quota issue then first thing comes to mind is food. I know it sucks but in an open market, as a property owner the landlord is free to choose the tenant and needs to be respected.