This is crazy. What percentage of Anthropic employees are non-American? And what happens with internal models that aren't exactly Fable/Mythos? Does this actually affect any employee access to SotA internal models?
Who's Who of AI
Marco
Why they matter
Practitioner with public evidence across Compute & infrastructure, Culture, work & education.
- AI signals
- 6 past 30d
- Sources
- 4 distinct domains
- Discussions
- 8 past 30d
- Latest signal
- 3d ago
Language and keyboard stuff at Google.
I like computers and Korean and computers-and-Korean and high school CS education.
Georgia Tech → 연세대학교 → 東京工業大学.
Regrettably no longer based in Tokyo :/
https://theoreticallygoodwithcomputers.com/
What they're sharing
Statement on the US government directive to suspend access to Fable 5 and Mythos 5 anthropic.com
Last Translation Benchmark arxiv.org
Chain of Thought Monitorability: A New and Fragile Opportunity for AI Safety arxiv.org
Epicure: Navigating the Emergent Geometry of Food Ingredient Embeddings arxiv.org
What's new in Claude Sonnet 5 simonwillison.net
View on Bluesky →
Alex L. Zhang | Speculative Programmatic Tool Calling alexzhang13.github.io
View on Bluesky →
Efficient Transformers with Dynamic Token Pooling aclanthology.org
FLEXITOKENS: Flexible Tokenization for Evolving Language Models aclanthology.org
MS MARCO: A Human Generated MAchine Reading COmprehension Dataset arxiv.org
Chessformer: A Unified Architecture for Chess Modeling arxiv.org
Needle 2 - The 14 MB Agentic LLM for Tiny Devices | Cactus cactuscompute.com
View on Bluesky →
Decoding GPT-Style Tokens mcognetta.github.io
View on Bluesky →
Articles & links
Marco reposted
Machine translation is not solved and it will take a while for it to be done arxiv.org/abs/2609.04173
AI Weekly's analysis
→
- Last Translation Benchmark ships peer-reviewed hard examples, each paired with handcrafted verification rules that flag concrete failure cases instead of assigning a metric score.
- LTBv1 covers submissions accepted before September 1st 2026, spans texts, images, audio and videos, and is designed to accept ongoing contributions.
- The authors argue MT is stuck: standard benchmarks are saturating, automatic metrics are gameable, and human evaluation lacks reproducibility and scalability.
Read full analysis →
View on Bluesky →
Yep, I was referring to things like aclanthology.org/2023.acl-lon... and aclanthology.org/2026.finding... but I agree. Reasoning in discrete text space is really nice for us as observers / auditors, but probably not optimal. arxiv.org/abs/2507.11473
AI Weekly's analysis
→
- A 41-author paper argues that reasoning models' human-language chains of thought are a monitorable safety signal that development choices could erode.
- The authors concede CoT monitoring is imperfect and misses some misbehavior, but urge investment in it alongside existing oversight methods.
- Signatories include Yoshua Bengio, Shane Legg, Geoffrey Irving, Neel Nanda, Ethan Perez, Evan Hubinger, Dan Hendrycks and Wojciech Zaremba, with Tomek Korbak leading.
Read full analysis →
Is there a link that others could use? Slightly related: arxiv.org/abs/2605.22391
Marco reposted
Notes (and a Pelican) on Claude Sonnet 5 - the new tokenizer makes it ~1.4x more expensive for English, ~1.33x more expensive for Spanish but roughly the same price for Simplified Mandarin simonwillison.net/2026/Jun/30/...
Marco reposted
Speculative tool calling from the author of RLMs, Alex Zhang yes, LLM text is streamed, so why not just predict what the next tool call is going to be and get started now before it finishes? boom. done. alexzhang13.github.io/blog/2026/sp...
Yep, I was referring to things like aclanthology.org/2023.acl-lon... and aclanthology.org/2026.finding... but I agree. Reasoning in discrete text space is really nice for us as observers / auditors, but probably not optimal. arxiv.org/abs/2507.11473
Yep, I was referring to things like aclanthology.org/2023.acl-lon... and aclanthology.org/2026.finding... but I agree. Reasoning in discrete text space is really nice for us as observers / auditors, but probably not optimal. arxiv.org/abs/2507.11473
I need to get on the team who made this first.
The Maia3 model source code was released today! github.com/CSSLab/maia3
Marco reposted
Saw the Needle2 45M LLM release posted on HN and thought it'd be great for open-source wearables, then I saw this... cactuscompute.com/needle
Marco reposted
Wrote a blog post about the token encoding format in GPT tokenizers. Have you ever inspected a tokenizer and seen tokens like оже or ĠнÑĥжно? These are tokens in a custom encoding format used for serialization of byte-level tokenizers. This post shows how to recover the…
Their own posts
Recent commentary
"We've just been doing light abuse so far. We are trying to kick it up now." -OpenAI
The Claude Code use case that has saved me the most time so far is just "get this slightly out of date ML repo working on my workstation".
Requesting that everyone call me Dr. Parameter-Efficient Korean Character-Level Language Modeling for now on. I'm honestly so disappointed in spammers these days. No excuse in the age of LLMs.
One funny thing about ChatGPT is that when I ask it some technical question and it starts doing the "searching the web" and then flashing through websites, it always does some weird ones like imbd. Idk what you are expecting to find there brother but you do you.
Why is there not a "I believe this paper is AI generated" button on paper review forms (that isn't shared with the author, but sent to the editors)? Ethical reviews are for ethical aspects of the paper, not ethical aspects of the writing process.
Their network
In Marco's orbit
Center = Marco. Left = members they follow (green edges). Right = members who follow them (blue edges). Top = mutual follows (orange edges, slightly larger). Drag any node to reposition; click to open that profile.
Are you Marco? Show it.
Add the Who’s Who of AI badge to your site or bio. It links back to this profile.
Markdown: [](https://aiweekly.co/whos-who/person/mcognetta-bsky-social)