iiinova — where your digital self keeps evolving

Lj Miranda @ljvmiranda921.itch.io · 27/09/2026

Rediscovering my love for electronics, first step is trying to deploy a model on an RPi 5: ljvmiranda921.github.io/notebook/202...

ljvmiranda921.github.io

Edge LM Devlog: Running a language model on a Raspberry Pi 5

A collection of notes, projects, and essays.

111

Lj Miranda @ljvmiranda921.itch.io · 03/09/2026

it's cool that 'teacher' and 'student' have the same number of letters so they align in my code

000

Reposted by Lj Miranda

Tom McCoy @rtommccoy.bsky.social · 20/08/2026

Since many are starting grad school soon, let me re-share my One Big Tip™️ for research! Research involves many skills - collaborating, writing, presenting, etc. But many of these skills can be unified under a single overarching ability: theory of mind Blog post link in reply

Illustration of the blog post's main argument, summarized as: "Theory of Mind as a Central Skill for Researchers: Research involves many skills.If each skill is viewed separately, each one takes a long time to learn. These skills can instead be connected via theory of mind – the ability to reason about the mental states of others. This allows you to transfer your abilities across areas, making it easier to gain new skills."

28019

Lj Miranda @ljvmiranda921.itch.io · 02/08/2026

Summer fun project! ljvmiranda921.itch.io/idle-frontier Idle Frontier is a one-click idle incremental game about the economics of training models. Annotate data, run experiments, chase grants and conference deadlines, and spend everything you earn on the next training run!

110

Lj Miranda @ljvmiranda921.itch.io · 06/07/2026

Yes, I still do ol' reliable NLP. Just released a maintenance version of calamanCy, a spaCy-based open source toolkit for Tagalog. Useful for dep. parsing, NER, tagging, etc. You can now install via `uv`! github.com/ljvmiranda92...

121

Reposted by Lj Miranda

Multilingual Representation Workshop @ EMNLP 2026 @mrl-workshop.bsky.social · 27/05/2026

📢 Call for Papers: 6th Multilingual Representation Learning Workshop at EMNLP in Budapest, Hungary! Join us and submit your works relating to multilingual NLP Speakers to be announced, so stay tuned! 👀 More info in the CFP: 🔗 sigtyp.github.io/ws2026-mrl.html

153

Lj Miranda @ljvmiranda921.itch.io · 20/08/2025

🇵🇭 One of my research interests is improving the state of Filipino NLP Happy to share that we're taking a major step towards this by introducing FilBench, an LLM benchmark for Filipino! Also accepted at EMNLP Main! 🎉 Learn more: huggingface.co/blog/filbench

huggingface.co

🇵🇭 FilBench - Can LLMs Understand and Generate Filipino?

140

Reposted by Lj Miranda

Ai2 @ai2.bsky.social · 28/07/2025

Ai2 is excited to be at #ACL2025 in Vienna, Austria this week. Come say hello, meet the team, and chat about the future of NLP. See you there! 🤝📚

093

Lj Miranda @ljvmiranda921.itch.io · 24/07/2025

I'll be at @aclmeeting.bsky.social‬ in Vienna! I'm going to present the ff first/co-first author works:

130

Reposted by Lj Miranda

SEACrowd @seacrowd.bsky.social · 16/05/2025

We’re thrilled that SEA-VL has been accepted to the ACL 2025 (Main)! Thank you to everyone who contributed to this project 🥳 Paper: arxiv.org/abs/2503.07920 Project: seacrowd.github.io/seavl-launch/ #ACL2025NLP #SEACrowd #ForSEABySEA

024

Reposted by Lj Miranda

Benjamin Minixhofer @bminixhofer.bsky.social · 02/04/2025

We created Approximate Likelihood Matching, a principled (and very effective) method for *cross-tokenizer distillation*! With ALM, you can create ensembles of models from different families, convert existing subword-level models to byte-level and a bunch more🧵

Image illustrating that ALM can enable Ensembling, Transfer to Bytes, and general Cross-Tokenizer Distillation.

12514

Reposted by Lj Miranda

Arduin Findeis @arduin.io · 17/03/2025

🕵🏻💬 Introducing Feedback Forensics: a new tool to investigate pairwise preference data. Feedback data is notoriously difficult to interpret and has many known issues – our app aims to help! Try it at app.feedbackforensics.com Three example use-cases 👇🧵

172

Reposted by Lj Miranda

Daniel van Strien @danielvanstrien.bsky.social · 13/03/2025

OLMo 2 0325 32B Preference Mixture: Solves AI alignment challenges through diverse preferences - Combines 7 datasets - Filters for instruction-following capability - Balances on-policy and off-policy prompts - Enabled successful DPO of OLMo-2-0325-32B model huggingface.co/datasets/all...

huggingface.co

allenai/olmo-2-0325-32b-preference-mix · Datasets at Hugging Face

We’re on a journey to advance and democratize artificial intelligence through open source and open science.

162

Reposted by Lj Miranda

Ai2 @ai2.bsky.social · 30/01/2025

Here is Tülu 3 405B 🐫 our open-source post-training model that surpasses the performance of DeepSeek-V3! It demonstrates that our recipe, which includes RVLR scales to 405B - with performance on par with GPT-4o, & surpassing prior open-weight post-trained models of the same size including Llama 3.1.

29221

Reposted by Lj Miranda

Kyle Lo @kylelo.bsky.social · 03/01/2025

kicking off 2025 with our OLMo 2 tech report while payin homage to the sequelest of sequels 🫡 🚗 2 OLMo 2 Furious 🔥 is everythin we learned since OLMo 1, with deep dives into: 🚖 stable pretrain recipe 🚔 lr anneal 🤝 data curricula 🤝 soups 🚘 tulu post-train recipe 🚜 compute infra setup 👇🧵

26917

Reposted by Lj Miranda

Tom Aarsen @tomaarsen.com · 19/12/2024

BERT is BACK! I joined a collaboration with AnswerAI and LightOn to bring you the next iteration of BERT. Introducing ModernBERT: 16x larger sequence length, better downstream performance (classification, retrieval), the fastest & most memory efficient encoder on the market. 🧵

1487

Reposted by Lj Miranda

Melissa Heikkilä @melissahei.bsky.social · 18/12/2024

New research reveals a worrying trend: AI's data practices risk concentrating power overwhelmingly in the hands of dominant technology companies. With analysis from @shaynelongpre.bsky.social @sarahooker.bsky.social @smw.bsky.social @giadapistilli.com www.technologyreview.com/2024/12/18/1...

technologyreview.com

This is where the data to build AI comes from

New findings show how the sources of data are concentrating power in the hands of the most powerful tech companies.

28530

Reposted by Lj Miranda

Jennifer Hu @jennhu.bsky.social · 07/12/2024

Stop by our #NeurIPS tutorial on Experimental Design & Analysis for AI Researchers! 📊 neurips.cc/virtual/2024/tutorial/99528 Are you an AI researcher interested in comparing models/methods? Then your conclusions rely on well-designed experiments. We'll cover best practices + case studies. 👇

neurips.cc

NeurIPS Tutorial Experimental Design and Analysis for AI ResearchersNeurIPS 2024

68614

Reposted by Lj Miranda

David Mimno @dmimno.bsky.social · 10/12/2024

We just updated the AI for Humanists guide to model selection to include Llama 3.3, and a recommended best cost/capability tradeoff, llama 3.1 8B. What have you tried, and what would you suggest? aiforhumanists.com/guides/models/

aiforhumanists.com

Models

The AI for Humanists project is developing resources to enable DH scholars to explore how large language models and AI technologies can be used in their research and teaching. Find an annotated biblio...

25419

Reposted by Lj Miranda

Kyle Lo @kylelo.bsky.social · 10/12/2024

the science of LMs should be fully open✨ today @akshitab.bsky.social @natolambert.bsky.social and I are giving our #neurips2024 tutorial on language model development. everything from data, training, adaptation. published or not, no secrets 🫡 tues, 12/10, 9:30am PT ☕️ neurips.cc/virtual/2024...

neurips.cc

NeurIPS Tutorial Opening the Language Model Pipeline: A Tutorial on Data Preparation, Model Training, and AdaptationNeurIPS 2024

514517

Reposted by Lj Miranda

Ian Magnusson @ianmagnusson.bsky.social · 10/12/2024

Come chat with me at #NeurIPS2024 and learn about how to use Paloma to evaluate perplexity over hundreds of domains! ✨We have stickers too✨

1214

Reposted by Lj Miranda

Grassroots Science @grassroots-science.bsky.social · 09/12/2024

⭐️ We're going to launch Grassroots Science, a year-long ambitious, massive-scale, fully open-source initiative aimed at developing multilingual LLMs aligned to diverse and inclusive human preferences in Feb 2025. 🌐 Check our website: grassroots.science. #NLProc #GrassrootsScience

grassroots.science

Grassroots Science

A global initiative focused on developing state-of-the-art multilingual language models through grassroots efforts.

175

Lj Miranda @ljvmiranda921.itch.io · 04/12/2024

We're releasing the largest Universal Dependencies (UD) treebank for Tagalog, UD-NewsCrawl! This dataset has been a long time coming, but glad to see this through: 15k+ sentences versus the previous ~150 sents from older Tagalog treebanks. 🤗 : huggingface.co/datasets/UD-... 📝 : Paper soon!

huggingface.co

UD-Filipino/UD_Tagalog-NewsCrawl · Datasets at Hugging Face

We’re on a journey to advance and democratize artificial intelligence through open source and open science.

1142

Reposted by Lj Miranda

Yoav Artzi @yoavartzi.com · 02/12/2024

I am seriously behind uploading Learning Machines videos, but I did want to get @jonathanberant.bsky.social's out sooner than later. It's not only a great talk, it also gives a remarkably broad overview and contextualization, so it's an excellent way to ramp up on post-training youtu.be/2AthqCX3h8U

youtu.be

Jonathan Berant (Tel Aviv University / Google) / Towards Robust Language Model Post-training

YouTube video by Yoav Artzi

15312

Lj Miranda @ljvmiranda921.itch.io · 26/11/2024

My favorite part about this release is that we were able to replicate our findings from the Tülu 3 post-training recipe here (e.g., on-policy preferences, RLVR) and found significant performance gains in our -DPO and -Instruct models! Find all artifacts here: huggingface.co/collections/...

huggingface.co

OLMo 2 - a allenai Collection

Artifacts for the second set of OLMo models.

030

Lj Miranda @ljvmiranda921.itch.io · 21/11/2024

Happy to be part of Tülu 3! Great effort to make the post-training stage open-source 😄 I worked on scaling our synthetic preference data (around 300k preference pairs for 70B) that led to performance gains when trained on using DPO.

170

Reposted by Lj Miranda

Ai2 @ai2.bsky.social · 21/11/2024

Meet Tülu 3, a set of state-of-the-art instruct models with fully open data, eval code, and training algorithms. We invented new methods for fine-tuning language models with RL and built upon best practices to scale synthetic instruction and preference data. Demo, GitHub, paper, and models 👇

211131

Reposted by Lj Miranda

Nathan Lambert @natolambert.bsky.social · 21/11/2024

I've spent the last two years scouring all available resources on RLHF specifically and post training broadly. Today, with the help of a totally cracked team, we bring you the fruits of that labor — Tülu 3, an entirely open frontier model post training recipe. We beat Llama 3.1 Instruct. Thread.

821343