Nathan Lambert

Nathan Lambert

@natolambert.bsky.social

A LLN - large language Nathan - (RL, RLHF, society, robotics), athlete, yogi, chef Writes http://interconnects.ai Prev Ai2/Olmo, HuggingFace, Berkeley, and normal places

14.9K followers278 following2.2K posts

Nathan Lambert
@natolambert.bsky.social

My first open model lecture, covering all the many pieces that build the ecosystem, and the open questions/debates determining where it goes next. How does distillation fit in? How do we manage risks? How do we see access to models changing in the future? Video: www.youtube.com/watch?v=79Pn...

youtube.comOpen Language Models: Overview, Open Source History, Distillation, Risks, & Chinese LabsFinally, I made the time to make some foundational educational content on open models. It's a very complex subject and it was a lot of fun to give a lecture that put together all aspects of my work…
Nathan Lambert
@natolambert.bsky.social

We got the Interconnects team together to debate hard questions and messy realities after a crazy, somewhat intimidating, week in the AI world. Interconnects Podcast #20: Kimi K3.1 and the new cyber reality YouTube: youtu.be/0xiq7RHVGRE Transcript / Podcast: www.interconnects.ai/p/podcast-ki...

youtu.beKimi K3.1 Pro Reactions and Living Through Hackmageddon | Interconnects PodcastPrograming update from Nathan: I'm going to start posting Interconnects content more directly on my @natolambert channel. It has better SEO and search performance, and it's easier than maintaining…
Nathan Lambert
@natolambert.bsky.social

TLDR on the kimi k3.1 pro benchmarks (just released) is that it looks like they figured out their RL, where the initial release was a very impressive base model with lighter post-training. China is very much keeping pace with capability gains.

View full thread
Nathan Lambert
@natolambert.bsky.social

Much to be learned in how the model stands up the vibe tests, but in the last few years the Moonshot team has rarely missed with the kimi models. K3.1 is also cheaper than K3 per token, but some benchmarks in the blog show it uses meaningfully more tokens per answer

Nathan Lambert
@natolambert.bsky.social

This, with all the cyber news, and rumors of a bigger GLM model coming soon, sets us up for an exciting rest of the year! The pro name makes me think Kimi will be moving to release flash model soon too, similar to DeepSeek and GLM.

Nathan Lambert
@natolambert.bsky.social

The Kimi team makes very clear statements about cyber risks and trying to pace the frontier of open-weight cyber capabilities and moderate it with respect to the closed frontier.

Nathan Lambert
@natolambert.bsky.social

Seems like cyber risk for AI may be more of an ecosystem issue than an open-weight models specific issue. If we can figure out the right balance here, I think there are a lot of people out there who want to support more openness in light of RSI and AI dynamics coming soon.

Nathan Lambert
@natolambert.bsky.social

Trying to use the models for things outside of coding / ml, like making slides or coherent plans or longish content design, makes me feel the AGI-is-going-to-take-a-while.

Nathan Lambert
@natolambert.bsky.social

In many ways, it feels like the epistemics of the models are getting way worse with increased RL -- they're super eager to solve, not eager to understand. Some older model, like GPT-4.5 and pre RL large models had a much better world model. I'm cautiously hoping Fable 5.5 helps solve some of this.

Nathan Lambert
@natolambert.bsky.social

A safe assumption when people are all commenting on a short Chinese policy document is that they don't know what it means and tried to parse a partial translation from claude, without reading all the way through. There are maybe 5-10 people who i trust to comment. I'm normally not one.

Nathan Lambert
@natolambert.bsky.social

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)

Nathan Lambert
@natolambert.bsky.social

Lots of ecosystem players (Nvidia, Huggingface) vague posting about a lot of releases in the near future. Rumors include thinky, muse spark, and Chinese players who haven’t dropped in a while. Fable 5.5 and whatever OpenAI is cooking expected too at any moment.

Nathan Lambert
@natolambert.bsky.social

Very timely appearance on the a16z podcast with Lisha Li, to try to understand how open models play into RSI dynamics, who captures value of open models, wrangling agent swarms, and why I'm building Trillium Labs: youtu.be/BKuK0M6oDTI?...

youtu.beCan Open Models Keep Up With OpenAI and Anthropic? | Nathan Lamberta16z Partner Lisha Li sits down with AI researcher and Trillium founder Nathan Lambert to ask how durable the gap really is between frontier AI labs and open models. Nathan argues that OpenAI and…
Nathan Lambert
@natolambert.bsky.social

We're hiring across many roles at Trillium Labs to build the open scientific ecosystem for LLMs. Most important hires right now (ideally with start dates soon): * Full time post-training researchers * Engineers with experience with popular open-source libraries...

Nathan Lambert
@natolambert.bsky.social

... * Mission-aligned generalists, non-profit ops, etc. * Student researchers I'm also personally interviewing for MATS right now, which will be very aligned research.

Nathan Lambert
@natolambert.bsky.social

If you announce a model's benchmarks without releasing it (weights, product, api, etc "later"), feels like almost a sure thing that someone releases a better model in your category before you get it out. ... Gemini Argon, Reflection Beam, who's next on this list?

Nathan Lambert
@natolambert.bsky.social

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

Nathan Lambert
@natolambert.bsky.social

The fact that all the AI news is so bizarre, kids vibe hacking with claude code, governments vibe governing, all massively increases the uncertainty of what's coming with AI. The meme-est AI timeline is also one of the scariest.

Nathan Lambert
@natolambert.bsky.social

Reflection joins the list of Nvidia and Thinking Machines who have tried to release their strongest models and come up behind Chinese counterparts. There's a lot of ways people will overthink this, but the clearest takeaway should be that the Chinese are very very good at building LLMs.

Nathan Lambert
@natolambert.bsky.social

I think pacing the frontier is good as an idea in principle but dead on arrival as something you can actually do. What will we, decide which benchmarks we can improve on? Who is involved? If we stop pushing "capabilities" we'll do a ton of research on swarms&efficiency which will bring other risks

Nathan Lambert
@natolambert.bsky.social

It distracts from how so much of AI risks is simply coming from diffusing what we have, and we should invest in preparedness, pushing the labs to be more careful, etc. Change my mind?

View full thread
Nathan Lambert
@natolambert.bsky.social

Trilliums are wildflowers that bloom briefly in the spring, before the forest canopies fill out. Though they are small, they lay the foundation for the cycles of growth and nourishment through the rest of the year.

Nathan Lambert
@natolambert.bsky.social

At Trillium Labs, the recipes will be the slow nutrients for the seasons and the model releases will be the blooms. Building an institution dedicated to this is needed because, much as nature’s trilliums are slow to expand and grow, the open-ecosystem needs time and dedicated resources to catch up.

Nathan Lambert
@natolambert.bsky.social

We're built around the theory of change that you need more eyes to solve hard technical problems. We have faith in the scientific methods and communities that humanity has built, and worry that AI is becoming too closed to utilize them.

Nathan Lambert
@natolambert.bsky.social

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.

Nathan Lambert
@natolambert.bsky.social

I'll have more to share soon, but it's important to have groups like this who are the independent thinkers at the cutting edge of the technology. Support like this has helped me take more personal career risk to do what I think is most urgently needed today.

Nathan Lambert
@natolambert.bsky.social

Love to see Google surprising people with Gemini 4. More labs at the frontier is wonderful for consumers (competition) and the world (reduction in concentration of power). Excited to see how it goes in real-world scenarios.

Nathan Lambert
@natolambert.bsky.social

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.

Nathan Lambert
@natolambert.bsky.social

Fairly easy actions I use to reduce phone addiction: 1. Never bring phone in the bedroom 2. If going out to a meal walking distance from home, leave phone behind 3. Black and white mode toggle on Home Screen 4. Silly one: move phone out of pockets I like to put my hands in

Nathan Lambert
@natolambert.bsky.social

Hearing AI regulation is in the air, especially with a misguided tone of “open dangerous, closed safe” so it felt timely to re-up this one. There are ways to make the AI industry safer without kneecapping transparency, education, and competition. www.interconnects.ai/p/banning-op...

interconnects.aiBanning Open Source AI Would Be A MistakeThis post was originally an op-ed co-authored with Kevin Xu of Interconnected for a general, non-technical audience.