iiinova — where your digital self keeps evolving

Nathan Lambert @natolambert.bsky.social · 15h

Guys this is like my superbowl I waited so long for US open weight models to be relevant again. Before today the US models were still behind Chinese models from as early as Feb. 2026. Still more work to do! This cut the gap by at least half, in time (say 3 months)

3524

Nathan Lambert @natolambert.bsky.social · 12/10/2026

Looks like Meta is back at the frontier with their latest "Muse Wonder" models. Will be interesting to see independent evaluations / user testing. research.meta.ai/blog/muse-wo...

3373

Nathan Lambert @natolambert.bsky.social · 10/10/2026

We've been monitoring a decline in HuggingFace downloads across the top open LLMs we track for interconnects. It looks like there was a slight change in counting of downloads, as we saw a ~30% download rate drop for many top models, but this is continuing over time. dashboard.interconnects.ai

4474

Nathan Lambert @natolambert.bsky.social · 09/10/2026

I’m more bought into AI becoming far more efficient, which is a massive lever on diffusion, than I am on AI models making massive breakthroughs that result in eliminating all the known weaknesses of LLMs. www.interconnects.ai/p/i-expect-r...

7472

Nathan Lambert @natolambert.bsky.social · 07/10/2026

In case you missed my hot take yesterday. www.interconnects.ai/p/the-cyber-...

817027

Nathan Lambert @natolambert.bsky.social · 01/10/2026

Excited to be in the 2026 early-career cohort of the Schmidt Sciences AI2050 fellowship. I'm thankful to have continued support of my work building and understanding the open ecosystem, so future AI is accessible, diffused widely, controlled by many, and safe.

2260

Nathan Lambert @natolambert.bsky.social · 30/09/2026

The latest Interconnects plot - showing the exponential growth of the open model inference economy. Plot is showing the daily tokens processed by the leading companies, based on public disclosures (for Together, Baseten, and Fireworks) and the OpenRouter API. I underestimated OpenRouter's growth.

2261

Nathan Lambert @natolambert.bsky.social · 21/09/2026

Summary on the state of open models, reflecting on my Congressional testimony on the topic. From: www.interconnects.ai/p/the-curren...

0124

Nathan Lambert @natolambert.bsky.social · 19/09/2026

A big problem with the AI forecasting discourse is that people ask “when will AI replace X job?” When in reality, AI will slowly replace a bunch of the sub-skills needed to do the job, but rarely all of them.

4265

Nathan Lambert @natolambert.bsky.social · 18/09/2026

With the latest external hack of OpenAI via Claude, closed models continue to be the tip of the iceberg on AI risks, not open models. They have been 1) much easier to get started with, 2) more capable & 3) shipped with leaky safeguards. Finetuning open models to specific attacks is harder.

74312

Nathan Lambert @natolambert.bsky.social · 15/09/2026

An basic idea in scaling RL: Can we allocate more compute to the harder problems? We did this: If your GRPO group has all wrong completions, sample more with probability P (~0.9) -- in search of more GRPO batches with nonzero gradient. It works! The paper: arxiv.org/abs/2609.13443

1598

Nathan Lambert @natolambert.bsky.social · 10/09/2026

Was spending too much time reading Tweets, so set out to write down why one resignation turned the embers of AI fear into a wildfire -- and importantly square which claims I think are valid or hyperbole in the resulting discourse. www.interconnects.ai/p/one-resign...

1153

Nathan Lambert @natolambert.bsky.social · 06/09/2026

Haven’t run much this year but sent 22 miles in the mountains to check in on the engine and we still got it. End of a great vaca.

0290

Nathan Lambert @natolambert.bsky.social · 25/08/2026

OOO for a bit, time to celebrate & rest :)

0261

Nathan Lambert @natolambert.bsky.social · 24/08/2026

Check out our latest open model data -- which models are mentioned in every arXiv ML paper since ChatGPT. A very fun weekend project with Codex! From idea is full dataset in <72hours. ~500K papers processed 😎 Plot: dashboard.interconnects.ai?axrange=all#...

5317

Nathan Lambert @natolambert.bsky.social · 17/08/2026

Personal milestone: 1000 true fans of my newsletter Interconnects! Hitting a very long term goal feels great. I’m happy to get to be an independent voice in AI. Cultivating a paid base helps me commit to that longer term, and scale Interconnects’ impact.

1460

Nathan Lambert @natolambert.bsky.social · 14/08/2026

Farewell Seattle! It's been such a wonderful life/career stage for me. Onto new adventures (and mountains).

1960

Nathan Lambert @natolambert.bsky.social · 10/08/2026

My book is 50% off on Manning until August 19th to celebrate the launch (code PBLambert). Here are 5 things you'll learn + a preamble on how you should use it! Here: www.interconnects.ai/p/5-useful-t...

3293

Nathan Lambert @natolambert.bsky.social · 07/08/2026

My summer project is done! A 20 video, free course on post-training to accompany my book is all on YouTube with slides open for modification & re-use. Plus, the book is 50% off right now with the code PBLambert on Manning to celebrate the launch. YouTube playlist: www.youtube.com/watch?v=jQPi...

1688

Nathan Lambert @natolambert.bsky.social · 03/08/2026

Introducing our Artifacts Hub and Adoption Dashboard Scaling our curation and measurement of the open ecosystem as we feel the acceleration of releases. Artifacts Hub: artifactshub.ai Adoption Dashboard: dashboard.interconnects.ai Explanation: www.interconnects.ai/p/introducin...

0132

Nathan Lambert @natolambert.bsky.social · 02/08/2026

Artifacts 23: Laguna S2.1, Inkling, & Kimi K3 show the utility of open models on the Pareto frontier. In this issue, a total of 24 models from june/july you should be aware of. www.interconnects.ai/p/latest-ope...

1210

Nathan Lambert @natolambert.bsky.social · 30/07/2026

Another new lecture! Lecture 11 is a tool-use/function calling/agentic 101. I almost skipped this one, as this chapter started as the only skill-specific topic in the book, but since writing it tool-use has only become more foundational to modern models.

2251

Nathan Lambert @natolambert.bsky.social · 30/07/2026

Frontier labs will be viable businesses by being able to integrate and optimize inference at lower cost/performance than most other models. They'll have a margin advantage on open models for the foreseeable future. This is one of those flex's imo and I doubt they're losing money.

1241

Nathan Lambert @natolambert.bsky.social · 28/07/2026

Lecture 10 of my course! Nominally on regularization in RL, so I discuss the evolving role of the KL penalty in RL, but also a set of nice RL papers that explain what RL helps models generalize better than SFT -- with theory supporting it.

1271

Nathan Lambert @natolambert.bsky.social · 28/07/2026

Kimi K3 with more likes than downloads on HuggingFace is definitely showing us a glimpse of the future on open models. It's way less about individual access, and more of a distributed platform layer for companies.

1452

Nathan Lambert @natolambert.bsky.social · 27/07/2026

Making talks with AI agents is awesome. I just told Fable to make a slide with real data on the KL distance from one of our reference Olmo 2 models and it made this with the wandb api (I edited text slightly).

2312

Nathan Lambert @natolambert.bsky.social · 24/07/2026

Insane numbers for opus 5, the power of faster iteration speed + scaled RL (Fable too big to RL as well, yet). And on safeguards "Based on our testing, we expect the classifiers to intervene around 85% less often than they do for Fable 5.". www.anthropic.com/news/claude-...

2789

Nathan Lambert @natolambert.bsky.social · 23/07/2026

If you're looking for the latest adoption data on open models in US v China v globally, we built a small dashboard with the big picture and per-org numbers. US's role is slowly growing, but still way behind China/Qwen. dashboard.interconnects.ai

2449

Nathan Lambert @natolambert.bsky.social · 22/07/2026

My book is the number 1 AI bestseller on Amazon. Is a success, even if that just lasts for a day :)! Thanks all for the support. rlhfbook.com

3650

Nathan Lambert @natolambert.bsky.social · 21/07/2026

My book, Reinforcement Learning from Human Feedback is done! This is the book I wish I had when learning to fine-tune, align, & now post-train models since ChatGPT. The resource has been built by me finding time to study and document the fundamentals on nights and weekends since 2024.

717521

Nathan Lambert @natolambert.bsky.social · 15/07/2026

Thinking Machines just released with a ~1T param, 41B active, apache-2 model Benchmarks are a clear step up from Nemotron Ultra (55B active), new best American model, and omni input. A bit behind GLM 5.2 on agentic benches, and Kimi K 2.6 on multi modal Super exciting!!

513916

Nathan Lambert @natolambert.bsky.social · 07/07/2026

Let’s goooo USA 🇺🇸🦅🏆

030

Nathan Lambert @natolambert.bsky.social · 28/06/2026

With everything going on, it gives me hope that there's such a diversity of companies building open models today. A lot of the story of open models unfolds under the shadow of the biggest frontier models. Lots of unearthed value.

1151

Nathan Lambert @natolambert.bsky.social · 25/06/2026

The goal with my rlhf book is to make the "home on the internet" for the next generation learning post-training. That's why I'm doing all formats (lectures, code, book, discord, model completions... & ofc blog of interconnects). A hub is more lasting than non-fiction writing. Join! rlhfbook.com

0399

Nathan Lambert @natolambert.bsky.social · 24/06/2026

"Knowledge wants to be free" is an unofficial motto for Interconnects and related projects like my book :) rlhfbook.com/course

0293

Nathan Lambert @natolambert.bsky.social · 22/06/2026

Excited to share a new open-source, RL recipe paper! TMax is the best openly available terminal-bench style training data, establishing the open frontier of small terminal agents with RL training. Many great insights into training in the work led by Hamish Ivison and Oscar Yin.

2376

Nathan Lambert @natolambert.bsky.social · 20/06/2026

Something I haven't advertised much is that I made a Discord to go with my RLHF book, launching in print in a few weeks. Trying to create the place for the next generation of folks trying to learn post-training to learn and have community. rlhfbook.com

1420

Nathan Lambert @natolambert.bsky.social · 16/06/2026

It's hard to pinpoint open-closed gap and so-on, but I trust the Arena team and just look where GLM 5.2 is on this. An MIT licensed, to be open weight model. At this point you could argue they have a better agent than Gemini does. That's a serious accomplishment.

212212

Nathan Lambert @natolambert.bsky.social · 15/06/2026

I launched 3 more videos in my post-training course! 1. Lecture 5: The rise of reasoning models 2. Lecture 6: DPO derivation, intuitions, and practice 3. A Q&A from readers on lectures 1-4 Course page: rlhfbook.com/course More soon!

2509

Nathan Lambert @natolambert.bsky.social · 10/06/2026

Why I think Anthropic's uneven safety policies with the release of Claude Fable 5 undermine the broader AI community's cohesion and accelerate us to more uncertainty and risk in AI's near-term evolution. www.interconnects.ai/p/claude-fab...

04010

Nathan Lambert @natolambert.bsky.social · 05/06/2026

Recording more post-training lectures with my favorite student

0150

Nathan Lambert @natolambert.bsky.social · 02/06/2026

I've attached the note I shared with the team and some fun photos from our time together. I'll keep cheering for Ai2 and am excited to see what you build next.

0160

Nathan Lambert @natolambert.bsky.social · 26/05/2026

Gemma 4 adoption numbers outpacing Qwen 3.5/3.6 for the same sized models is a big shift in the international balance of influence via open models.

3588

Nathan Lambert @natolambert.bsky.social · 26/05/2026

The geopolitical angle of the AI race is only going to keep accelerating. More domestic talent travel restrictions from the Chinese government. I expect more and more things like this, AI is getting embedded in the core of existing power structures. www.bloomberg.com/news/article...

0325

Nathan Lambert @natolambert.bsky.social · 17/05/2026

Being out of SF has lowered my information proximity but with the big upside of giving me space to cultivate my own beliefs and values around ai. We need more people zagging in AI, the monoculture just helps the incumbents win at this point.

1473