Michael (Ai4Privacy) (@MikeDoesDo) profile picture

Large Language Model Expert Master in Artificial Intelligence

Joined August 2016
Photo by Michael (Ai4Privacy) (@MikeDoesDo)Photo by Michael (Ai4Privacy) (@MikeDoesDo)Photo by Michael (Ai4Privacy) (@MikeDoesDo)Photo by Michael (Ai4Privacy) (@MikeDoesDo)Photo by Michael (Ai4Privacy) (@MikeDoesDo)Photo by Michael (Ai4Privacy) (@MikeDoesDo)Photo by Michael (Ai4Privacy) (@MikeDoesDo)Photo by Michael (Ai4Privacy) (@MikeDoesDo)Photo by Michael (Ai4Privacy) (@MikeDoesDo)Photo by Michael (Ai4Privacy) (@MikeDoesDo)
Michael (Ai4Privacy) retweeted
It was a pleasure to talk about human evals last Monday and share a preview of one of our upcoming LLM benchmarks! Especially since I was sharing the stage with my two friends @flschottmann and @MikeDoesDo.
2
4
127
Michael (Ai4Privacy) retweeted
30M+ human judgements across image, video, 3D and world models, all available on our HF. This week's highlights: -Gemini 4 Argon #1 in Text-to-3D -FLUX 3 Max #1 in physics -FIBO 2.0 #2 among open-weight image models
2
7
21
416,115
Michael (Ai4Privacy) retweeted
Apply to speak at @EthCC, and join us for @TheAgora_Event by @KaikoData next April in Cannes!
Speaker applications for EthCC[10] are officially open. For nearly a decade, the EthCC stage has hosted the industry's most important conversations. In 2027, join the ranks of industry leaders and share your insights at the iconic Palais des Festivals in Cannes. Have a perspective worth bringing to the stage? We want to hear it.
6
2
23
382
Michael (Ai4Privacy) retweeted
Replying to @imjaredz
This is an amazing 10x improvement on an already super small model. You crushed it, literally.
1
1
9
262
Michael (Ai4Privacy) retweeted
New Benchmark: Image-to-3D with 307K+ human votes across 14 models. Hi3D v3.0 leads the leaderboard, with Meshy v7.1 and Tripo H3.1 among the top ones. We evaluated geometry and visual faithfulness across 357 input images. Full results on benchmark.ai
2
1
4
259
Michael (Ai4Privacy) retweeted
I'll be in Singapore next week for token2049 for some ultrarare time away from the charts hope to see some of you friends there
3
1
17
543
Michael (Ai4Privacy) retweeted
Rampart, our PII removal model, has cracked the first screen of the top trending models across any category on Huggingface, on the same tier as GLM 5.2 / Deepseek! If building systems at fast pace at huge scale is interesting to you, reach out
7
9
2
180
21,304
Michael (Ai4Privacy) retweeted
🇺🇸🇺🇸 @ndstudio
Rampart from the @ndstudio @WhiteHouse is number one trending token classification model on HF. Very cool to see public organizations starting to own and build their weights instead of renting them from an API provider! huggingface.co/models?pipeli…
5
3
109
16,522
Michael (Ai4Privacy) retweeted
Amazing.
Today, we are releasing Rampart: a 14.7MB machine learning model designed to protect citizens’ privacy by redacting personal information directly in your browser before it gets sent to any server
32
100
5
2,571
372,844
Michael (Ai4Privacy) retweeted
FULL INTERVIEW: Engineers Edward Coristine (@as400495) and Tai Groot (@taigrr) just released an ML model called Rampart for the National Design Studio. It's a local-first, open-source AI privacy model that runs entirely in the browser. They joined TBPN to explain the model and why the National Design Studio launched it. 0:25 – National Design Studio launches Rampart 1:40 – Why Rampart is open source 2:40 – How NDS trained the model 3:55 – Use cases for the model 5:25 – How NDS thinks about side quests 6:25 – Team size, workflow, and why they work seven days a week 8:00 – Recruiting top talent to the National Design Studio
Tomorrow, we're releasing an ML model for on-device PII redaction, which is far smaller than existing models, but is surprisingly effective. It is a first iteration. Will be live at 12:30!
13
23
5
198
119,918
Michael (Ai4Privacy) retweeted
🎯
FULL INTERVIEW: Engineers Edward Coristine (@as400495) and Tai Groot (@taigrr) just released an ML model called Rampart for the National Design Studio. It's a local-first, open-source AI privacy model that runs entirely in the browser. They joined TBPN to explain the model and why the National Design Studio launched it. 0:25 – National Design Studio launches Rampart 1:40 – Why Rampart is open source 2:40 – How NDS trained the model 3:55 – Use cases for the model 5:25 – How NDS thinks about side quests 6:25 – Team size, workflow, and why they work seven days a week 8:00 – Recruiting top talent to the National Design Studio
4
6
1
154
23,345
Michael (Ai4Privacy) retweeted
Better than the frontier model at a tiny portion of the side Government writing good tech wasn’t in the deck of cards until now
8
15
1
181
28,791
Michael (Ai4Privacy) retweeted
Rampart is trained on synthetic data, open source datasets, and government FOIA releases, including OCR-processed documents to improve performance on noisy data
5
6
273
28,546
Michael (Ai4Privacy) retweeted
The raw text you enter into a chatbot might include names, addresses, or financial details. Rampart detects and replaces sensitive terms with stable placeholders like [NAME_1], and sends the cleaned version instead. This way, the model still understands your request but your data doesn't have to leave the device
6
24
3
462
49,591
Michael (Ai4Privacy) retweeted
GM Rampart (evals from a subset of OpenPII 1.5M). Our first iteration, will improve from here
Tomorrow, we're releasing an ML model for on-device PII redaction, which is far smaller than existing models, but is surprisingly effective. It is a first iteration. Will be live at 12:30!
4
7
1
83
13,393
Michael (Ai4Privacy) retweeted
National Design Studio announces ‘Rampart’, a 14.7MB machine learning model that redacts personal information directly in browser before being sent to any server to protects citizens’ privacy. Edward Coristine (@as400495) & Tai Groot (@taigrr) explain: nitter.cf/i/broadcasts/1dKrPPRmg…
8
10
79
17,204
Michael (Ai4Privacy) retweeted
Introducing Rampart, an open source model to protect the privacy of citizens’ personal information, built by the talented software team at @ndstudio. Watch the recorded live stream below with @as400495 @taigrr 🇺🇸
Today, we are releasing Rampart: a 14.7MB machine learning model designed to protect citizens’ privacy by redacting personal information directly in your browser before it gets sent to any server
15
23
1
337
39,213
Michael (Ai4Privacy) retweeted
This is our first iteration of Rampart. We built it because we wanted Americans to have better PII removal on-device along with other privacy preserving measures. If you are motivated by the work of building better technology to serve Americans, consider joining National Design Studio.
4
10
310
27,045
Michael (Ai4Privacy) retweeted
Today, we are releasing Rampart: a 14.7MB machine learning model designed to protect citizens’ privacy by redacting personal information directly in your browser before it gets sent to any server
279
867
171
6,976
2,224,409