Christopher Akiki

data and research at @lichess

[Christopher Akiki @christopher]

@kalomaze Would be interesting to compare to this: https://t.co/JbbnVDPasi

All tokens of all BPE tokenizers on the HF hub. (Though we do need to rehydrate with new tokenizers)

[Christopher Akiki @christopher]

@charles_irl AND PROCESSING DATA

[Christopher Akiki @christopher]

@MeganRisdal ❤️

[Christopher Akiki @christopher]

The @huggingface version of our chess position evaluations dataset just crossed 1 billion rows. That's 1 billion @stockfishchess evaluations from games played on @lichess!

[Christopher Akiki @christopher]

- https://t.co/lmfeLNUc00
- https://t.co/PS6j1F3YVw
- https://t.co/KBhQJtAf3R

[Christopher Akiki @christopher]

Interesting release by @jure Leskovec's group: RT-J (86M "relational transformer" ) and THE JOIN, a web-crawled corpus of 639 relational databases with 6K forecasting tasks. (data and models on @huggingface)

[Christopher Akiki @christopher]

@natolambert @trillium_labs Is remote a dealbreaker?

[Christopher Akiki @christopher]

@kaggle The CTFs organized by the folks who went on to found @dreadnode ! cc @kenyeung_tech @moo_hax @bradpalmtree

[Christopher Akiki @christopher]

@irombie New album cover

[Christopher Akiki @christopher]

@tulkenss @LucileSaulnier Do you have any guesses?

[Christopher Akiki @christopher]

Can't get Schmidhubered if I Schmidhuber myself.

[Stanford NLP Group @stanfordnlp]

Everyone wants those AI scientists! But how possible is it for an LLM to generate novel research ideas rather than just optimizing hyperparameters (broadly construed)?

Building directly on the once novel work of @SchmidhuberAI!

[Christopher Akiki @christopher]

@sarahookr found it 🫵

[Christopher Akiki @christopher]

@stevenstrogatz @EdwRobinson @nntaleb https://t.co/mhDLqVgmB7

It's also a field of study!

[Christopher Akiki @christopher]

@GiadaPistilli 🐱🐈🐈‍⬛😻😽😸😺

[Christopher Akiki @christopher]

@danqi_chen @AdithyaNLP @jeff_cheng_77 Any chance this was recorded?

[Christopher Akiki @christopher]

@TZahavy @AatuSelkee @ericmalmi @Xidong_Feng Very cool work!

[Marco🍞 @marco_computers]

🚨 [Token][ization] Paper Alert 🚨

Tokenization is a wildly understudied area of language modeling despite it having effects across all of NLP. Over the past ~8 months, 32 (!) tokenizer researchers put together the most comprehensive survey of the field. Check it out!

[Christopher Akiki @christopher]

@durham_uni cc @cgeorgiaw

[Christopher Akiki @christopher]

@leland_mcinnes @lu__jasper This looks insanely cool. Is this happening in a branch you could link to? Or in stealth mode?

PS: UMAP never stopped being cool.