iiinova — where your digital self keeps evolving

Kyle Lo @kylelo.bsky.social · 06/10/2026

I'm at #colm2026, Mon-Thurs Supporting two papers from @ai2.bsky.social days 🐟Olmo Hybrid, 7B hybrid model 🐠Cracks in the Foundation, long context recipes Hoping to have fun chats w folks about scaling LM data & evals, ☺️ unbiased on pretrain, posttrain, all the 🚂s

1170

Kyle Lo @kylelo.bsky.social · 01/05/2026

during in Olmo 3 we thought long context is just finding good data nope! model architecture matters & it's hard to recover if mess it up led by @abertsch.bsky.social, we release many pretrain runs w/ small arch changes and show huge long context performance diffs

05811

Kyle Lo @kylelo.bsky.social · 28/03/2026

Today I'm saying farewell to @ai2.bsky.social. I'm so proud of our team & grateful to have shared fully-open Olmo, Dolma, olmOCR, Molmo, etc with the world I know the team is more committed than ever to advancing open-source & open-science. Forever rooting for my dear friends 🫶

3541

Kyle Lo @kylelo.bsky.social · 05/03/2026

our new Olmo Hybrid model combines attention with linear RNN layers 🍣training efficiency is crazy good. the model reaches same MMLU score as Olmo 3 in 50% of the tokens. also see this in many other tasks as always: weights, data, ckpts, training code, etc. all fully open

1354

Kyle Lo @kylelo.bsky.social · 13/02/2026

one of my favorite topics is dealing with data constraints! what if your proposed mix is 30% code but you don't have enough code? we can repeat our data until we hit target proportions, but too much is risky we view data mixing as (data) constrained optimization

130

Kyle Lo @kylelo.bsky.social · 11/02/2026

literally all the time 😮‍💨 this was yesterday

030

Kyle Lo @kylelo.bsky.social · 04/02/2026

our open model proving out specialized rag LMs over scientific literature has been published in nature ✌🏻 congrats to our lead @akariasai.bsky.social & team of students and Ai2 researchers/engineers www.nature.com/articles/s41...

24310

Kyle Lo @kylelo.bsky.social · 27/01/2026

The 5th Generation, Evaluation, and Metrics (GEM) Workshop will be at #ACL2026! Call for papers is out. Topics include: 🐟 LMs as evaluators 🐠 Living benchmarks 🍣 Eval with humans and more New for 2026: Opinion & Statement Papers! Full CFP: gem-workshop.com/call-for-pap...

0227

Kyle Lo @kylelo.bsky.social · 18/01/2026

lucky to chat w sen. patty murray about olmo & importance of fully open AI

2501

Kyle Lo @kylelo.bsky.social · 07/01/2026

just in case it wasn’t clear which room this is

120

Kyle Lo @kylelo.bsky.social · 12/12/2025

paper has: 🐟 more on our eval ideology 🦈 more baselines 🍣 more about RL Zero etc we picked final model (internally called moonlit surfer 🌛🏄) not just on bench scores but good vibes 🥰

110

Kyle Lo @kylelo.bsky.social · 01/12/2025

I'll be at #NeurIPS2025 from Tues-Sat! Come say hi 👋 if you wanna chat about 🦈 olmo 3 stories 🐟 pretraining data & evals 🍣 midtraining shouldnt exist 🐠 model specialization 🐡 AI for education 🍥 tabletop games

1192

Kyle Lo @kylelo.bsky.social · 21/11/2025

yess!! sry bout the x-axis, still thinkin how to make figure clearer it's exactly what you're saying -- each point refers to a stage of development. our release has data+ckpts+evals for all stages we use (figure) and wanted to show how it compares to other models which typically only few stages

020

Kyle Lo @kylelo.bsky.social · 20/11/2025

🍕Finally, we all know midtraining is an exciting time to get a ton of performance boost But team organization to sustain consistent model improvements (without burnout) is important! We have explorers "own" target capabilities & centralized assessment team run "integration tests"

110

Kyle Lo @kylelo.bsky.social · 20/11/2025

🍨Data quality signals matter but also how you use them! Traditional ways of using data quality is to threshold: Define a cutoff and take all the documents above that threshold. But why not sample *proportional* to data quality? We use Quality-Aware Upsampling to do exactly this

130

Kyle Lo @kylelo.bsky.social · 20/11/2025

🍣Data mixing is a little too powerful It's easy to learn "optimal" mixes that oversample from certain pockets heavily. eg, STEM docs are valuable for climbing MMLU & but you don't have infinite STEM docs We approach mixing as Token Constrained Optimization over diverse evals

140

Kyle Lo @kylelo.bsky.social · 20/11/2025

🦈Invest in your experimental design! We create evals better suited for different compute scales, with our "easy" set of tasks+metrics able to support very small scale experiments before switching to our "main" set of evals, on which smaller models are below noise floor

170

Kyle Lo @kylelo.bsky.social · 20/11/2025

we released Olmo 3! lot of exciting stuff but wanna focus on: 🐟Olmo 3 32B Base, the best fully-open base model to-date, near Qwen 2.5 & Gemma 3 on diverse evals 🐠Olmo 3 32B Think, first fully-open reasoning model approaching Qwen 3 levels 🐡12 training datasets corresp to different staged training

1417

Kyle Lo @kylelo.bsky.social · 05/11/2025

why intern at Ai2? 🐟interns own major parts of our model development, sometimes even leading whole projects 🐡we're committed to open science & actively help our interns publish their work reach out if u wanna build open language models together 🤝 links 👇

2278

Kyle Lo @kylelo.bsky.social · 22/10/2025

woah guess VLMs for OCR the hottest research topic this week😆 since the first olmOCR, we've been.. 🔥training our VLM using RLVR with binary unit test rewards🔥 it's incredibly effective & unit test creation easy to scale w synthetic data pipelines check it out at olmocr.allen.ai

0213

Kyle Lo @kylelo.bsky.social · 11/10/2025

bye #colm2025 big fan of the montreal bagels 🥯 hot take I like them better than

0120

Kyle Lo @kylelo.bsky.social · 09/10/2025

lol so much love for prepost-postpre training

000

Kyle Lo @kylelo.bsky.social · 09/10/2025

any other fans of pre-pretraining?

120

Kyle Lo @kylelo.bsky.social · 09/10/2025

come say hi at posters this morning for OLMo 2 and fluid benchmarking posters 👋 and dont miss @valentinhofmann.bsky.social's talk in morning #colm2025 @ai2.bsky.social vry proud of my gifs

270

Kyle Lo @kylelo.bsky.social · 08/10/2025

@josephc.bsky.social @mariaa.bsky.social and I are at poster #21 findings from large scale survey of 800 researchers on how they use LMs in their research #colm2025

0163

Kyle Lo @kylelo.bsky.social · 06/10/2025

flyin to #colm2025 along w bunch of the @ai2.bsky.social team come chat w me about pretraining horror stories, data & evals, what we're cookin for next olmo, etc made a 🔥 poster for thursday sess, come say hi

0111

Kyle Lo @kylelo.bsky.social · 06/10/2025

5 am airport for the only direct flight from seattle to montreal #colm2025

1120

Kyle Lo @kylelo.bsky.social · 17/09/2025

LM benchmark design requires 3 decisions, how to: 🐟 select test cases 🐠 score LM on each test 🦈 aggregate scores to estimate perf fluid benchmarking is simple: 🍣 find max informative test cases 🍥 estimate 'ability', not simple avg perf why care? turn ur grey noisy benchmarks to red ones!

052

Kyle Lo @kylelo.bsky.social · 20/08/2025

looks like the preprint has been updated to include a disclaimer that this was a class project & intentionally provocatively written 😐

150

Kyle Lo @kylelo.bsky.social · 08/08/2025

⚠️ AI-generated content may be inaccurate. Verify important information independently.

040

Kyle Lo @kylelo.bsky.social · 01/07/2025

only took few days to descend into madness

0120

Kyle Lo @kylelo.bsky.social · 26/06/2025

back from copenhagen & berkeley travels, now moving into new @ai2.bsky.social office!

0160

Kyle Lo @kylelo.bsky.social · 21/06/2025

thx for organizing! great to meet NLP folks & consume fancy bread 🥖🍞🥐

0210

Kyle Lo @kylelo.bsky.social · 19/06/2025

the benchmark works based on thousands of "unit tests" so instead of fuzzy matching between a model-generated table with a gold reference table, we define Pass/Fail tests like "the cell to the left of the cell containing 0.001 should contain 1.96"

120

Kyle Lo @kylelo.bsky.social · 13/06/2025

we won honorable mention for Best Paper at #CVPR2025 🏆 for Molmo & Pixmo, showing the value of high-quality data for VLMs! recalling when we released same time as Llama 3.2 😆 huge kudos to Matt Deitke, Chris Clark & Ani Kembhavi for their leadership on this project! @cvprconference.bsky.social

0343

Kyle Lo @kylelo.bsky.social · 12/06/2025

google down, guess ill go smell flowers or sthn 🤷‍♂️

040

Kyle Lo @kylelo.bsky.social · 12/06/2025

excited to see this release of 1M public domain & CC zero books, digitized and OCR'd! 👏 big win for open data, congrats to the authors! arxiv.org/abs/2506.08300

0365

Kyle Lo @kylelo.bsky.social · 29/05/2025

looks like same group got an AI generated paper accepted to ACL 😅 www.intology.ai/blog/zochi-acl

470

Kyle Lo @kylelo.bsky.social · 23/05/2025

hilarious deep research UX w/ the agent trace it's like "i found relevant content in <journal|conf|arxiv> paper" but the links provided all go to the publisher homepage instead of the actual paper lolol whyy 🤦‍♂️

180

Kyle Lo @kylelo.bsky.social · 23/05/2025

nice article thx for sharing! enjoyed fig about surveying bad baselines

010

Kyle Lo @kylelo.bsky.social · 15/05/2025

unfortunately not 😮‍💨 it's disabled for me too; i am wondering if the call was incorrect - the real for making sure all openreview profiles exist was actually the abstract deadline, not full submission deadline. ill try emailing PCs

100

Kyle Lo @kylelo.bsky.social · 12/05/2025

@neuripsconf.bsky.social it seems the call for papers neurips.cc/Conferences/... says author list should be finalized by May 15th, but on OpenReview itself, author list needs to be finalized by May 11th can pls clarify, thx! 🙏

230

Kyle Lo @kylelo.bsky.social · 10/05/2025

haerfest in wallingford is excellent! 😋

1291

Kyle Lo @kylelo.bsky.social · 28/04/2025

it's today! im speaking at the Data for Foundation Models workshop #ICLR2025 today Mon 4/28, after lunch at 2pm SGT, Hall 4 #4 get ready for 🐟 lessons we've learned from OLMo 1 to OLMo 2 🐡 unpublished methods & results 🐠 practical advice for academics & smaller teams 🐈‍⬛ cats on slides!

1141

Kyle Lo @kylelo.bsky.social · 25/04/2025

starting now in Hall 3 poster 306! come chat w me @muennighoff.bsky.social & @soldaini.net about our MoE w fully open data, weights, code, and more! complete w iOS app to run it completely on device 😆 poster is so pretty 😭 and don’t miss oral session later too! #ICLR2025

190

Kyle Lo @kylelo.bsky.social · 24/04/2025

excited to be at #ICLR2025 w/ the @ai2.bsky.social OLMo team presenting some of our work (below) come chat w/ us to learn more! we'd love to collaborate & we're also hiring :D

140

Kyle Lo @kylelo.bsky.social · 23/04/2025

some food pics from short stop in japan 🇯🇵 now off to singapore 🇸🇬 for #ICLR2025 & some chicken rice / roti prata 🤤

1150

Kyle Lo @kylelo.bsky.social · 25/03/2025

2. after recording, they have option to see playback & redo recording before submitting

110

Kyle Lo @kylelo.bsky.social · 25/03/2025

sry late to reply! i couldn't find the original implementation, but here's an example I mocked up to give you a sense of the workflow: 1. annotators see something like this & have to press Start to record

110