DALL-E 2 has a secret language
twitter.com
DALL-E 2 has a secret language
1–10 of 118 posts
Re: DALL-E 2 has a secret language
#2I bet it's just a form of copy protection.
Re: DALL-E 2 has a secret language
#3I bet it's just a form of copy protection.
Re: DALL-E 2 has a secret language
#4That's reminiscent of small children making up their own words for things. Those words are stable in that you can converse with the child using those words.
Re: DALL-E 2 has a secret language
#5In short: DALLE-2 generates apparent gibberish for text in some circumstances, but feeding the gibberish back in gets recognized and you can tease out the meaning of words in this unknown language.
Re: DALL-E 2 has a secret language
#6Link to the 5 page paper, for those that don't like twitter threads:
https://giannisdaras.github.io/publications/Discovering_the_...
Re: DALL-E 2 has a secret language
#7Possibly related: In 2017 AI bots formed a derived shorthand that allowed them to communicate faster: https://www.facebook.com/dhruv.batra.dbatra/posts/1943791229...
> While the idea of AI agents inventing their own language may sound alarming/unexpected to people outside the field, it is a well-established sub-field of AI, with publications dating back decades.
> Simply put, agents in environments attempting to solve a task will often find unintuitive ways to maximize reward.
Re: DALL-E 2 has a secret language
#8It’s wild to see the discoveries being made in ML research. Like most of these ‘discoveries,’ it makes a fair amount of sense after thinking about it. Of course it’s not just going to spit out random noise for random input, it’s been trained to generate realistic looking images.
But I think it is an interesting discovery because I don’t think anyone could have predicted this.
One of my favorite examples is the classification model that will identify an apple with a sticker on it that says “pear” as a pear—it makes sense, but is still surprising when you first see it.
Re: DALL-E 2 has a secret language
#9Was DALL-E 2 trained on captions from multiple languages? If so, this makes a lot of sense. Somewhere early in the model the words "bird", "vogel", "oiseau" and "pájaro" have to be mapped to the same concept. And "Apoploe vesrreaitais" happens to map to the same concept. Or maybe "Apoploe vesrreaitais" is rather the tokenization of that concept, since it also appears in the output. So in a sense DALL-E is using an internal language to make sense of our world.
Re: DALL-E 2 has a secret language
#10Interestingly Google detects these words as Greek. I know they are nonsensical and not actually Greek but I'm wondering if any Greek speakers might be able to provide some insights. Are these gibberish words close to meaningful words? (clear shot in the dark here) Maybe a linguist could find more meaning?