Christopher Akiki (@christopher) header
Christopher Akiki (@christopher) profile picture

Christopher Akiki

@christopher

data, research, and emotes at @lichess.

 Leipzig, Germany
hf.co/christopher
Joined July 2009
  • Tweets1.7K
  • Following214
  • Followers5.2K
250Media
Photo by Christopher Akiki (@christopher)Photo by Christopher Akiki (@christopher)Photo by Christopher Akiki (@christopher)Photo by Christopher Akiki (@christopher)Photo by Christopher Akiki (@christopher)Photo by Christopher Akiki (@christopher)Photo by Christopher Akiki (@christopher)Photo by Christopher Akiki (@christopher)Photo by Christopher Akiki (@christopher)Photo by Christopher Akiki (@christopher)
Le Chonk's Tokenizer. LeChonkenizer?
model_type: BPE vocab_size: 204800 num_special_slots: 1000 num_named_special: 28 num_filler_special: 972 num_bpe_tokens: 203800 num_merges: 423969 merges_per_token: 2.08 ignore_merges: True byte_fallback: False normalizer: None pre_tokenizer: ['Split', 'ByteLevel'] post_processor: ByteLevel decoder: ByteLevel regex: [^\r\n\p{L}\p{N}]?[\p{Lu}\p{Lt}\p{Lm}\p{Lo}\p{M}]*[\p{Ll}\p{Lm}\p{Lo}\p{M}]+|[^\r\n\p{L}\p{N}]?[\p{Lu}\p{Lt}\p{Lm}\p{Lo}\p{M}]+[\p{Ll}\p{Lm}\p{Lo}\p{M}]*|\p{N}{1,3}(?=(?:\p{N}{3})*(?!\p{N}))|\p{N}{1,2}| ?[^\s\p{L}\p{N}]+[\r\n/]*|\s*[\r\n]+|\s+(?!\S)|\s+ special_tokens: [(0, '<unk>'), (1, '<s>'), (2, '</s>'), (3, '[INST]'), (4, '[/INST]'), (5, '[AVAILABLE_TOOLS]'), (6, '[/AVAILABLE_TOOLS]'), (7, '[TOOL_RESULTS]'), (8, '[/TOOL_RESULTS]'), (9, '[TOOL_CALLS]'), (10, '[IMG]'), (11, '<pad>'), (12, '[IMG_BREAK]'), (13, '[IMG_END]'), (14, '[PREFIX]'), (15, '[MIDDLE]'), (16, '[SUFFIX]'), (17, '[SYSTEM_PROMPT]'), (18, '[/SYSTEM_PROMPT]'), (19, '[TOOL_CONTENT]'), (24, '[AUDIO]'), (25,

ALT model_type: BPE vocab_size: 204800 num_special_slots: 1000 num_named_special: 28 num_filler_special: 972 num_bpe_tokens: 203800 num_merges: 423969 merges_per_token: 2.08 ignore_merges: True byte_fallback: False normalizer: None pre_tokenizer: ['Split', 'ByteLevel'] post_processor: ByteLevel decoder: ByteLevel regex: [^\r\n\p{L}\p{N}]?[\p{Lu}\p{Lt}\p{Lm}\p{Lo}\p{M}]*[\p{Ll}\p{Lm}\p{Lo}\p{M}]+|[^\r\n\p{L}\p{N}]?[\p{Lu}\p{Lt}\p{Lm}\p{Lo}\p{M}]+[\p{Ll}\p{Lm}\p{Lo}\p{M}]*|\p{N}{1,3}(?=(?:\p{N}{3})*(?!\p{N}))|\p{N}{1,2}| ?[^\s\p{L}\p{N}]+[\r\n/]*|\s*[\r\n]+|\s+(?!\S)|\s+ special_tokens: [(0, '<unk>'), (1, '<s>'), (2, '</s>'), (3, '[INST]'), (4, '[/INST]'), (5, '[AVAILABLE_TOOLS]'), (6, '[/AVAILABLE_TOOLS]'), (7, '[TOOL_RESULTS]'), (8, '[/TOOL_RESULTS]'), (9, '[TOOL_CALLS]'), (10, '[IMG]'), (11, '<pad>'), (12, '[IMG_BREAK]'), (13, '[IMG_END]'), (14, '[PREFIX]'), (15, '[MIDDLE]'), (16, '[SUFFIX]'), (17, '[SYSTEM_PROMPT]'), (18, '[/SYSTEM_PROMPT]'), (19, '[TOOL_CONTENT]'), (24, '[AUDIO]'), (25,

Mistral AI (@MistralAI) profile pictureMistral AI
@MistralAIOct 29
Today, we release the weights of a new and improved Le Chonk, advancing the open frontier for agentic coding and cybersecurity. 💻 Built to build: Le Chonk ranks #1 on @cognition’s FrontierCode among open models, and sets a new benchmark for Pareto efficiency on @artificialanalysis. 🛡 Built to defend: State-of-the-art in cyberdefense, put through the paces by the world's leading cyber vendors, including @CrowdStrike, @cloudflare, @rubrik, @zscaler, and others. Built for sovereignty: Make it yours. Download the weights on @huggingface , customize it with Mistral Forge, and deploy it wherever you need it. Read more from our team and customers 👇:
Image from Mistral AI (@MistralAI)
1
10
360
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!
Screenshot of the dataset viewer on hugging face of the lichess evaluations dataset.

ALT Screenshot of the dataset viewer on hugging face of the lichess evaluations dataset.

4
3
18
4,258
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)
Image from Christopher Akiki (@christopher)
3
1
1
19
1,158
- arxiv.org/abs/2610.17867v3 - huggingface.co/datasets/stan… - huggingface.co/stanford-star…
1
3
160
Can't get Schmidhubered if I Schmidhuber myself.
Stanford NLP Group (@stanfordnlp) profile pictureStanford NLP Group
@stanfordnlpOct 6
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!
1
4
264
Christopher Akiki retweeted
Marco🍞 (@marco_computers) profile picture
Marco🍞
@marco_computersSep 30
🚨 [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!
Image from Marco🍞 (@marco_computers)
6
83
24
346
40,972
Christopher Akiki retweeted
Maxime Labonne (@maximelabonne) profile picture
Maxime Labonne
@maximelabonneSep 21
ChessLFM went full circle and is now on the home page of @lichess Super cool to see, Lichess is where I collected all the seed data. Based on my limited chess skills, I never thought I'd have an article there.
Image from Maxime Labonne (@maximelabonne)
4
1
56
3,634
Love this new slop option on linkedin. They seem to at least be self-aware.
linkedin report message menu: post hidden in feed. You'll Hide posts by Jamie Seems like AI slop (underlined) Report post

ALT linkedin report message menu: post hidden in feed. You'll Hide posts by Jamie Seems like AI slop (underlined) Report post

5
588
Very interesting looking viz fail.
odd looking graph. everything is bundled. colorful edges. black background.

ALT odd looking graph. everything is bundled. colorful edges. black background.

2
12
809
t-SNE on SigLIP embeddings of 1,360 @OReillyMedia cover animals. The penguin cluster is my favorite.
t-SNE plot of animal engravings.

ALT t-SNE plot of animal engravings.

Christopher Akiki (@christopher) profile pictureChristopher Akiki@christopherJul 16
O'Reilly animals sorted by orientation using PCA.
Image quilt of O'Reilly media animals sorted by direction.

ALT Image quilt of O'Reilly media animals sorted by direction.

3
16
126
7,223
O'Reilly animals sorted by orientation using PCA.
Image quilt of O'Reilly media animals sorted by direction.

ALT Image quilt of O'Reilly media animals sorted by direction.

Christopher Akiki (@christopher) profile pictureChristopher Akiki@christopherJul 13
Image quilt of all @OReillyMedia book covers sorted by color.
List of all O'Reilly books, sorted by hue.

ALT List of all O'Reilly books, sorted by hue.

15
189
13
2,432
197,003
Image quilt of all @OReillyMedia book covers sorted by color.
List of all O'Reilly books, sorted by hue.

ALT List of all O'Reilly books, sorted by hue.

Christopher Akiki (@christopher) profile pictureChristopher Akiki@christopherJul 11
An image quilt of all @OReillyMedia animals.
small multiple photo of 1360 oreilly animals

ALT small multiple photo of 1360 oreilly animals

1
10
1
179
206,040
An image quilt of all @OReillyMedia animals.
small multiple photo of 1360 oreilly animals

ALT small multiple photo of 1360 oreilly animals

1
4
1
32
8,488
Christopher Akiki retweeted
Ian Johnson 🔬🤖 (@enjalot) profile picture
Ian Johnson 🔬🤖
@enjalotJul 11
single 5090 running parametric umap on 8 million points. getting ready to push to 16 million then 150 million 🌌
Image from Ian Johnson 🔬🤖 (@enjalot)
2
4
19
1,141
The McManifold (2216 burger recipes as 146-dimensional ingredient vectors)
small multiples umap plots of the burger dataset.

ALT small multiples umap plots of the burger dataset.

Nature Portfolio (@NaturePortfolio) profile pictureNature Portfolio
@NaturePortfolioJul 3
Using burgers as a model system, a generative AI system rediscovers the classic Big Mac without explicit supervision and generates novel burgers optimized for deliciousness, sustainability, or nutrition, according to a paper in npj Science of Food. go.nature.com/3SU7JDQ
This is figure 2, which shows rediscovering classic burger and discovering delicious burgers.

ALT This is figure 2, which shows rediscovering classic burger and discovering delicious burgers.

4
17
1,473
Some fun looking UMAP tomfoolery.
UMAP plot.

ALT UMAP plot.

1
9
554
4 years ago today, BLOOM finished training.
BigScience Large Model Training (@BigScienceLLM) profile pictureBigScience Large Model Training@BigScienceLLM28 Jun 2022
▓▓▓▓▓▓▓▓▓▓▓▓▓▓▓ 100%
2
1
26
3,338
10M Bohemian eigenvalues of all 19x19 {-1, 1} companion matrices.
rainbow colormap scatterplot of a fractal structure. Circular, symmetric.

ALT rainbow colormap scatterplot of a fractal structure. Circular, symmetric.

6
287
work in progress: 1 million synthetic personas from @NVIDIAAI's Nemotron-Personas-USA dataset.
scatter plot of 1 million personas from https://huggingface.co/datasets/nvidia/Nemotron-Personas-USA

ALT scatter plot of 1 million personas from https://huggingface.co/datasets/nvidia/Nemotron-Personas-USA

1
9
553
Image from Christopher Akiki (@christopher)
4
157
Load more