In OpenAI's o200k_base tokenizer, " strawberry" is a single token (101830): a small demo of why letter-counting is awkward for LLMs

cognenza@lemmy.world to

LocalLLaMA@sh.itjust.worksEnglish · 4 days ago

Ran this with tiktoken: enc = tiktoken.get_encoding("o200k_base"); enc.encode("How many r's are in strawberry?") → [5299, 1991, 428, 885, 553, 306, 101830, 30]. Without the leading space it splits differently: "strawberry" → st | raw | berry. Spelled out, "s t r a w b e r r y" is 10 tokens, one per letter. Tokenization is only part of the story: CharBench (AAAI) finds word length and the actual count matter more for counting tasks. Still, it’s a neat way to see that the model gets IDs, not characters.

You must log in or register to comment.

Chat

LocalLLaMA@sh.itjust.works

localllama@sh.itjust.works

Create a post

You are not logged in. However you can subscribe from another Fediverse account, for example Lemmy or Mastodon. To do this, paste the following into the search field of your instance: !localllama@sh.itjust.works

Welcome to LocalLLaMA! Here we discuss running and developing machine learning models at home. Lets explore cutting edge open source neural network technology together.

Get support from the community! Ask questions, share prompts, discuss benchmarks, get hyped at the latest and greatest model releases! Enjoy talking about our awesome hobby.

As ambassadors of the self-hosting machine learning community, we strive to support each other and share our enthusiasm in a positive constructive way.

Rules:

Rule 1 - No harassment or personal character attacks of community members. I.E no namecalling, no generalizing entire groups of people that make up our community, no baseless personal insults.

Rule 2 - No comparing artificial intelligence/machine learning models to cryptocurrency. I.E no comparing the usefulness of models to that of NFTs, no comparing the resource usage required to train a model is anything close to maintaining a blockchain/ mining for crypto, no implying its just a fad/bubble that will leave people with nothing of value when it burst.

Rule 3 - No comparing artificial intelligence/machine learning to simple text prediction algorithms. I.E statements such as "llms are basically just simple text predictions like what your phone keyboard autocorrect uses, and they’re still using the same algorithms since <over 10 years ago>.

Rule 4 - No implying that models are devoid of purpose or potential for enriching peoples lives.

Visibility: Public

This community can be federated to other instances and be posted/commented in by their users.

674 users / day
1.16K users / week
1.16K users / month
1.17K users / 6 months
1 local subscriber
5.51K subscribers
60 Posts
494 Comments
Modlog