Home Artificial Intelligence Translate with ChatGPT Translation Prompt General translation Domain and Robustness Limitations of this study Conclusion

Translate with ChatGPT Translation Prompt General translation Domain and Robustness Limitations of this study Conclusion

60
Translate with ChatGPT
Translation Prompt
General translation
Domain and Robustness
Limitations of this study
Conclusion

Image from Pixabay.

ChatGPT is a chatbot developed by OpenAI. It relies on instructGPT: It has been trained to follow and answer instructions, or so-called “prompts,” written by users.

ChatGPT demonstrates impressive abilities in providing coherent and relevant detailed answers to user prompts. It seems to resembling summarization, query answering, language generation, and .

Nonetheless, because it is a really recent system, ChatGPT to match its NLP performance with previous work.

Towards that direction, Tencent AI published a preliminary study on ChatGPT’s ability to translate:

Is ChatGPT A Good Translator? A Preliminary Study by Wenxiang Jiao, Wenxuan Wang, Jen-tse Huang, Xing Wang, and Zhaopeng Tu (Tencent AI)

The principal objective of this study is to judge ChatGPT for translating text into English since most of its training data is in English. Note: Indeed, ChatGPT relies on instructGPT, as mentioned within the blog post. InstructGPT is GPT-3 fine-tuned with prompts “mostly in English” (Ouyang et al., 2022). Furthermore, 93% of GPT-3’s pre-training data is English (Brown et al., 2020).

In addition they evaluate translation into other languages which are much less represented in its training data, resembling Japanese and Romanian, and thus tougher.

In this text, I’ll analyze and explain their principal findings, especially to spotlight what seems to work and what doesn’t when using ChatGPT as a machine translation system.

When coping with generative language models, one of the necessary steps is prompt design.

We want to seek out an appropriate natural language formulation to question the model given our goal task. Here we would like ChatGPT to translate a sentence in a source language, denoted “[SRC],” right into a goal language, denoted “[TGT].”

To search out good prompts, Tencent AI directly asked ChatGPT to offer 10 prompts, with the next prompt:

Provide ten concise prompts or templates that could make you translate.

ChatGPT returned as expected 10 prompts, but with only . They finally determine to try only the next 3 that are essentially the most representative of the ten prompts initially returned by ChatGPT:

  • Prompt 1: Translate these sentences from [SRC] to [TGT]:
  • Prompt 2: Answer with no quotes. What do these sentences mean in [TGT]?
  • Prompt 3: Please provide the [TGT] translation for these sentences:

They evaluated each certainly one of these prompts on a Chinese-to-English translation task ([SRC]=Chinese, [TGT]=English), and obtained the next results:

Results by Wenxiang Jiao, Wenxuan Wang, Jen-tse Huang, Xing Wang, and Zhaopeng Tu (Tencent AI)

BLEU, chrF++, and TER are 3 automatic metrics for evaluating machine translation quality. With BLEU and chrF++, higher scores are higher. With TER, lower scores are higher.

Based on the scores obtained with these 3 metrics, they found that Prompt 3 performs the perfect. Prompt 2 seems also higher than Prompt 1, despite the fact that chrF++ scores look similar.

That is interesting because Prompt 1 mentions the source language but the opposite two prompts don’t. Yet, Prompt 1 underperforms. .

That is impressive but in addition counter-intuitive. We could have expected ChatGPT to be more accurate due to the precision of the source language in its prompts. For human translators, knowing the source language is critical.

Currently, there isn’t a good explanation for why ChatGPT yields lower scores when indicating the source language. We will assume that ChatGPT can routinely infer the source language from the user input. If so, providing the source language shouldn’t have any impact, as an alternative of the negative impact observed in Tencent AI results.

Now that we have now found a very good prompt, we are able to evaluate ChatGPT against state-of-the-art machine translation systems.

Tencent AI selected online systems for comparisons: Google Translate, DeepL, and their very own online system, Tencent TranSmart.

The outcomes are as follows:

Results by Wenxiang Jiao, Wenxuan Wang, Jen-tse Huang, Xing Wang, and Zhaopeng Tu (Tencent AI)

The three online systems perform similarly and appear to perform higher than ChatGPT, despite the fact that the authors don’t report on statistical significant testing to make sure that that the differences are really significant.

Yet, I discovered these results impressive. Being based on instructGPT, we are able to assume that ChatGPT is principally , but seems in a position to to generate English translations.

If we could fine-tune ChatGPT for Chinese-to-English, we might definitely obtain a translation of a much higher quality.

Within the paper, Tecent AI also reports on similar differences for all translation directions between English, Chinese, German, and Romanian.

Table by Wenxiang Jiao, Wenxuan Wang, Jen-tse Huang, Xing Wang, and Zhaopeng Tu (Tencent AI)

Again, the performances (in BLEU) are impressive. Even for translation directions that don’t involve English, resembling German-to-Chinese, ChatGPT can generate translations. In line with BLEU, online systems remain higher, as expected since they’re trained for this task. ChatGPT isn’t!

Results involving Romanian are quite different. As an illustration, the BLEU rating is nearly 50% lower for ChatGPT in comparison with the net systems. This difference might be statistically significant.

The authors propose a proof. Romanian is a language for which far fewer resources, e.g, Romanian text on the Web, can be found than for German and Chinese. ChatGPT can have seen during its training to accurately model them.

I might agree with this assumption, but it surely ought to be confirmed with more experiments involving other languages with similar amounts of resources, resembling Croatian or Polish.

They carried out further experiments to judge the performance of ChatGPT in translating texts in a and (posted on , normally very noisy with grammatical errors).

Table by Wenxiang Jiao, Wenxuan Wang, Jen-tse Huang, Xing Wang, and Zhaopeng Tu (Tencent AI)

Surprisingly, the performance of ChatGPT stays near online systems for translating biomedical texts from German-to-English, in keeping with BLEU.

ChatGPT doesn’t appear to be negatively impacted by the very specific terms utilized in biomedical texts.

ChatGPT. That is impressive, but less surprising. We will assume that ChatGPT has , while the net systems training data used for comparison are frequently heavily curated, and thus somewhat less robust to errors (grammatical, semantic, etc.).

This task is far more difficult for ChatGPT when translating into languages distant from English, resembling Japanese as shown by the outcomes on WMT20 Rob2, as expected.

The authors acknowledge of their study that more experiments with more language pairs are mandatory to raised assess ChatGPT’s translation quality.

This assessment ought to be performed with human evaluation relatively than with automatic metrics which are often inaccurate, especially when the scores of the systems compared are very close.

The .

In my view, the impact of the prompt could possibly be also further investigated. The authors selected a really original way by letting ChatGPT itself suggest prompts. But is a chicken and egg problem. The prompt itself used to get prompts for machine translation can have a robust impact on all the next experiments performed on this study. Previous work on prompt designing for machine translation tried very diverse and handcrafted prompts.

ChatGPT is .

From this preliminary study, we are able to already conclude that ChatGPT could be good, and possibly even higher than standard online systems, at translating text for which the interpretation is anticipated to have the characteristics of ChatGPT’s training data, as an example, noisy user-generated texts in English.

Yet, as expected, ChatGPT remains to be behind more standard machine systems for translating into languages aside from English, especially distant or low-resource languages, resembling Japanese or Romanian.

60 COMMENTS

  1. … [Trackback]

    […] Read More: bardai.ai/artificial-intelligence/translate-with-chatgpttranslation-promptgeneral-translationdomain-and-robustnesslimitations-of-this-studyconclusion-2/ […]

  2. … [Trackback]

    […] Here you will find 57088 additional Information on that Topic: bardai.ai/artificial-intelligence/translate-with-chatgpttranslation-promptgeneral-translationdomain-and-robustnesslimitations-of-this-studyconclusion-2/ […]

  3. Hey I know this is off topic but I was wondering if you
    knew of any widgets I could add to my blog that automatically tweet my newest twitter updates.
    I’ve been looking for a plug-in like this for quite some time and was hoping maybe
    you would have some experience with something like this.
    Please let me know if you run into anything. I truly enjoy reading your blog and I look forward
    to your new updates.

  4. A fascinating discussion is worth comment. I think that you ought to write more about this topic,
    it might not be a taboo subject but generally people do not talk about such topics.
    To the next! Best wishes!!

  5. I have been exploring for a little for any high quality articles or blog posts in this
    kind of house . Exploring in Yahoo I finally stumbled upon this site.
    Reading this info So i’m happy to exhibit that I’ve an incredibly just right uncanny feeling I found out just what I needed.
    I so much without a doubt will make sure to do not overlook this website and give it
    a look on a constant basis.

  6. Hey! I know this is kind of off topic but I was wondering
    which blog platform are you using for this site? I’m getting tired of WordPress because I’ve
    had issues with hackers and I’m looking at alternatives for another platform.
    I would be great if you could point me in the direction of a good platform.

  7. I really like what you guys are usually up too. Such clever work
    and exposure! Keep up the great works guys I’ve
    incorporated you guys to my blogroll.

  8. Thanks for some other excellent post. Where else may just anybody get
    that type of info in such an ideal manner of writing?
    I’ve a presentation subsequent week, and I’m at the
    search for such information.

  9. Good day! This post couldn’t be written any better!
    Reading this post reminds me of my previous room mate! He always kept talking about this.

    I will forward this page to him. Pretty sure he will have a good read.
    Thank you for sharing!

  10. Do you mind if I quote a few of your posts as long as I provide
    credit and sources back to your blog? My website is in the very
    same area of interest as yours and my users would truly benefit from a lot of the information you provide here.

    Please let me know if this ok with you. Thanks a lot!

  11. Hello there, just became alert to your blog through Google,
    and found that it is really informative. I am gonna watch out for brussels.
    I will be grateful if you continue this in future. Lots of people will be benefited from your writing.

    Cheers!

  12. I am not sure where you’re getting your information, but great topic.
    I needs to spend some time learning much more or understanding more.
    Thanks for fantastic info I was looking for this info for my mission.

  13. Just wish to say yоսr article is as surprising.
    The ϲlearness in yojr post is sіmply excellent and i can assume you’re an expert on this subject.
    Well with your permiѕsion let me to grab yoir feed t᧐ keеp up to date with forthcoming pоst.

    Тhanks a million and please keep սp the enjoyable work.

  14. Hi there, just became alert to your blog through Google,
    and found that it’s truly informative. I’m going to watch out for brussels.
    I’ll be grateful if you continue this in future. Many people will be benefited from
    your writing. Cheers!

  15. Wow, marvelous blog format! How lengthy have you ever been blogging for?
    you made blogging glance easy. The whole glance of your
    site is fantastic, let alone the content material!

  16. This design is wicked! You definitely know how to keep a
    reader amused. Between your wit and your videos, I was almost moved to start my own blog (well, almost…HaHa!) Great job.
    I really enjoyed what you had to say, and more than that, how
    you presented it. Too cool!

  17. This is very interesting, You’re a very skilled blogger.
    I have joined your rss feed and look forward to seeking more of your excellent post.

    Also, I have shared your web site in my social networks!

    my page … Puff Wow

  18. I loved as much as you’ll receive carried out right here.
    The sketch is attractive, your authored subject matter stylish.
    nonetheless, you command get got an nervousness over that you wish be delivering the following.
    unwell unquestionably come further formerly again since exactly the same nearly very often inside case you shield this increase.

  19. Thanks for one’s marvelous posting! I actually enjoyed reading it,
    you are a great author.I will make sure to bookmark your blog and will often come back someday.

    I want to encourage continue your great posts,
    have a nice evening!

  20. The other day, while I was at work, my cousin stole my iphone and tested to see if it can survive a 30
    foot drop, just so she can be a youtube sensation. My iPad is
    now destroyed and she has 83 views. I know this is completely off topic but I had to share it with someone!

  21. Hello, i read your blog from time to time and i
    own a similar one and i was just curious if you get a lot of spam feedback?

    If so how do you protect against it, any plugin or anything you can suggest?

    I get so much lately it’s driving me crazy so any assistance is very much appreciated.

  22. Greate article. Keep posting such kind of info on your blog.
    Im really impressed by it.
    Hey there, You have performed a fantastic job.
    I will definitely digg it and personally recommend to my friends.
    I am confident they will be benefited from this website.

  23. Hi there, i read your blog occasionally and i own a similar one and i was just curious if you get a lot
    of spam comments? If so how do you protect against it, any plugin or anything
    you can suggest? I get so much lately it’s driving
    me mad so any support is very much appreciated.

LEAVE A REPLY

Please enter your comment!
Please enter your name here