For the unfamiliar: https://www.youtube.com/watch?v=TJYOkZz6Dck

(CW: gore, subtitles)

In Germany, Mein Kampf is banned except for educational purposes, eg in history class.

Strictly speaking this is incorrect, although the situation is somewhat complicated. There are laws that can be and were used to limit its redistribution (mainly the rule against anti-constitutional propaganda), but there are dissenting judgements saying original prints from before the end of WW2 cannot fall under this, since they are pre-constitutional. One particular reprint from 2018 has been classified as “liable to corrupt the young”, but to my knowledge this only means it cannot be publicly advertised.

What is interesting though is how distribution and reprinting was prevented historically, which is copyright. As Hitlers legal heir the state of Bavaria held the copyright until it expired in 2015 and simply didn’t grant license to anything except versions with scholarly commentary. But technically since then anybody can print and distribute new copies of the book. If this violates any law will then be determined on a case-by-case basis after the fact.

pet open source projects that no one else ever seems to contribute to, not […] software that holds up civilization

SamePicture.jpeg

deleted by creator

If you rise anywhere above lever 5 or so, the difficulty ratchets up so much it makes the main quest nearly impossible to complete.

Didn’t Oblivion already have the difficulty slider? You could just adjust that, no?

I know level scaling is a big topic in the industry, but for me, the way it’s implemented nearly ruins what is otherwise a mostly great game.

Two of the first RPGs I played were Gothic and Gothic II which released approximately alongside Morrowind and Oblivion, and they just had no dynamic level scaling at all, so I don’t really see the appeal either. A tiny Mole Rat being roughly the same challenge as a big bad Orc just breaks immersion. If you were to meet the latter in early game it would just curb stomp you, which provided an immersive way of gating content and a real sense of achievement when you came back later with better armour and weapons to finally defeat the enemy who gave you so many problems earlier. Basically the same experience you had with Death Claws in Fallout New Vegas when compared to Fallout 3 - they aren’t just a set piece, they are a real challenge.

The games had their own problems, for example the fighting system sucked, and I’m told the English translation was so bad the games just flopped in the Anglosphere, putting them squarely in the Eurojank category of games. But creating a real sense of progression and an immersive world were certainly not amongst their weaknesses.

deleted by creator

pandoc.org is probably what you are looking for, but you might have to create a custom reader/writer or find one on the internet.

If you have a mouse wheel try pressing it down while hovering over a link. Should open in new tab.

a neural network with a series of layers (W in this case would be a single layer)

I understood this differently. W is a whole model, not a single layer of a model. W is a layer of the Transformer architecture, not of a model. So it is a single feed forward or attention model, which is a layer in the Transformer. As the paper says, a LoRA:

injects trainable rank decomposition matrices into each layer of the Transformer architecture

It basically learns shifting the output of each Transformer layer. But the original Transformer stays intact, which is the whole point, as it lets you quickly train a LoRA when you need this extra bias, and you can switch to another for a different task easily, without re-training your Transformer. So if the source of the bias you want to get rid off is already in these original models in the Transformer, you are just fighting fire with fire.

Which is a good approach for specific situations, but not for general ones. In the context of OP you would need one LoRA for fighting it sexualising Asian women, then you would need another one for the next bias you find, and before you know it you have hundreds and your output quality has degraded irrecoverably.

Yeah but that’s my point, right?

That

1. you do not “replace data until your desired objective”.
2. the original model stays intact (the W in the picture you embedded).

Meaning that when you change or remove the LoRA (A and B), the same types of biases will just resurface from the original model (W). Hence “less biased” W being the preferable solution, where possible.

Don’t get me wrong, LoRAs seem quite interesting, they just don’t seem like a good general approach to fighting model bias.

First, there is no thing as a “de-biased” training set, only sets with whatever target series of biases you define for them to reflect.

Yes, I obviously meant “de-biased” by definition of whoever makes the set. Didn’t think it worth mentioning, as it seems self evident. But again, in concrete terms regarding the OP this just means not having your dataset skewed towards sexualised depictions of certain groups.

1. either you replace data until your desired objective, which will reduce the model’s quality for any of the alternatives

[…]
For reference, LoRAs are a sledgehammer approach to apply the first way.

The paper introducing LoRA seems to disagree (emphasis mine):

We propose Low-Rank Adaptation, or LoRA, which freezes the pre-trained model weights and injects trainable rank decomposition matrices into each layer of the Transformer architecture, greatly reducing the number of trainable parameters for downstream tasks.

There is no data replaced, the model is not changed at all. In fact if I’m not misunderstanding it adds an additional neural network on top of the pre-trained one, i.e. it’s adding data instead of replacing any. Fighting bias with bias if you will.

And I think this is relevant to a discussion of all models, as reproduction of training set biases is something common to all neural networks.

“Inclusive models” would need to be larger.

[citation needed]

To my understanding the problem is that the models reproduce biases in the training material, not model size. Alignment is currently a manual process after the initial unsupervised learning phase, often done by click-workers (Reinforcement Learning from Human Feedback, RLHF), and aimed at coaxing the model towards more “politically correct” outputs; But ultimately at that time the damage is already done since the bias is encoded in the model weights and will resurface in the outputs just randomly or if you “jailbreak” enough.

In the context of the OP, if your training material has a high volume of sexualised depictions of Asian women the model will reproduce that in its outputs. Which is also the argument the article makes. So what you need for more inclusive models is essentially a de-biased training set designed with that specific purpose in mind.

I’m glad to be corrected here, especially if you have any sources to look at.

Ubuntu ∈ Linux distributions ⇒ Linux can do that

*coughs in snap*

Ok you actually made me curious so I threw in a quick search, turns out they aren’t opposed to technology or electricity per se at all, it’s a much more philosophical stance about being connected to / dependant on the outside world. And it’s indeed different for every community/parish to what degree it is allowed, some more conservative groups still don’t use it all.

Some Amish, though not all, also accept the use of solar panels to generate energy to charge batteries, power an electric fence for livestock, or heat water. Donald Kraybill has called this form of electricity tapping into “God’s grid”. […]

The Amish are not against use of electric power and acknowledge its usefulness. They seek to remain off the public grid in order to prevent worldly influences from entering the home, and as a symbolic means of remaining separate from the world.

At the same time, they see value in limited use of electric power, and thus generate it by various means, making use of diesel generators, batteries, inverters, and solar panels, among other technologies.

https://amishamerica.com/do-amish-use-electricity/

There are still some extremely conservative orders (also called “low” orders or “old” orders) of Amish that still do not allow the use of batteries either in the home or as safety lights on buggies. […]

As history marched on, and new inventions were discovered and marketed, lines had to be drawn as to what was, and what was not acceptable within the Amish church. The acceptance of electricity within the home was one of these lines. By the 1920’s most Amish churches had agreed on a ban for Amish church members being allowed to connect to the electrical grid. […]

Today, most Amish churches forbid the use of public electricity because it is seen as a “connection to the world” but batteries allow many Amish families and businesses a limited connection to power and the ability to run such items as calculators, alarm clocks, cash registers, drills, electric fences, and even cell phones without discipline form their local Bishop.

http://www.amishcountryalmanac.com/2014/02/batteries-and-amish.html

Now please don’t ask me how they rationalise buying generators and fuel as not being dependant on the outside world, because I don’t have the slightest clue.

Gasoline powered generators apparently, and then the big batteries charged on those can charge the phones or power whatever they need. But again, this is different from community to community as far as I understood it.

Based on a documentary I saw recently they seem to be OK with battery powered devices and use them extensively on their farms. But powerlines are considered the devil or something. And autmobiles as well, unless you get a chauffer. Not sure how much this differs between communities though, doesn’t seem like there is a central authority for all Amish.

If I’m understanding your question right, kind of. Pandoc is only for document conversion though, no spreadsheets, presentations, etc. But at that it can convert between a lot of formats. And git can be used to version and share those documents.

Nothing, I’d just like a nice GUI around it.

Well like uzay said, basically just an office GUI that allows me to import/export into a lot of formats and automates document versioning away.