ex-Google/ initial Borg team/ creator of GCL, golang.org/x/text, CUE, etc.
Joined July 2015
- Tweets 317
- Following 158
- Followers 1,302
- Likes 797
3 Photos and videos
Make GitHub Actions, and the rest of your CI/CD and workflow stack, more powerful with type safety, validation, reusability, and policy enforcement using CUE!
How? Easy - check out this hands-on demo from our community member @rawkode:
youtube.com/watch?v=MFtQAhIq…
1
6
11
2,799
📣 Meet the "Shift-Left" Editor.
Join us next Tuesday, Nov 18, at 16:00 UTC for a hands-on demo of the new CUE LSP. Get real-time validation, error-checking and language features as you type - right inside your editor.
Learn more call in details: github.com/cue-lang/cue/disc…
4
9
875
Marcel van Lohuizen retweeted
10 Nov 2025
This paper from Tsinghua University and Shanghai Jiao Tong University received perfect scores (6, 6, 6, 6) at NeurIPS 2025!
It aims to answer a key question: Does reinforcement learning really make large language models better reasoners?
The authors study Reinforcement Learning with Verifiable Rewards (RLVR) and find that while it improves accuracy for small k, it doesn’t create new reasoning patterns—meaning the base model still determines the upper limit of reasoning ability.
Across six RLVR variants, performance gains plateau, suggesting that current RL setups mainly refine reasoning rather than reinvent it.
Interestingly, it’s distillation, not RL, that shows genuine signs of emergent reasoning.
This research points to the next frontier for truly self-improving large language models.
Does Reinforcement Learning Really Incentivize Reasoning Capacity in LLMs Beyond the Base Model?
Paper: arxiv.org/abs/2504.13837
Page: limit-of-RLVR.github.io
Our report: mp.weixin.qq.com/s/2-GDxs8j1…
📬 #PapersAccepted by Jiqizhixin
8 Nov 2025
Thrilled that our paper received the only perfect score at NeurIPS this year.
Huge thanks to my collaborators and the reviewers. See you in San Diego!
limit-of-rlvr.github.io
papercopilot.com/statistics/…
credit to @papercopilot
28
274
1,772
396,954
📣 We have just released #cuelang v0.15.0 - introducing the CUE LSP, two new language experiments, and much, much more 🎉
github.com/cue-lang/cue/rele…
3
9
1,144
3 Nov 2025
Looking forward to this!
Final call! 📣
Join us with Kelsey Hightower tomorrow, Tuesday, Nov 4, at 17:00 UTC to explore the future of configuration with our Configuration Control Plane, and turn configuration from a source of risk into a source of reliability!
Register here: luma.com/ysloofz3
1
184
Marcel van Lohuizen retweeted
30 Oct 2025
Today, I'm proud to officially announce CUE Labs!
I co-founded CUE Labs with @mpvl_ to build a Configuration Control Plane, solving config chaos with CUE. Our work also provides the stewardship for the @cue_lang open source project.
Read more: cue.dev/blog/announcing-cue-…
3
7
433
30 Oct 2025
I'm proud to officially announce CUE Labs!
I co-founded CUE Labs with @_myitcv to build a Configuration Control Plane, solving config chaos with CUE. Our work also provides the stewardship for the @cue_lang open source project.
Read our launch post: cue.dev/blog/announcing-cue-…
1
2
7
927
🔈#KubeCon London opens in 2 days! Will you be there? Let us know: forms.gle/L62aiNHc1R8tjgPs6 - we’ll save you a cool CUE cap. We’re bringing lots of updates & demos: If you use GitHub Actions, Azure, GitLab, Argo, or K8s, our CUE content can simplify your workflow. Can’t wait! 🚀
1
3
636
24 Sep 2023
It was a privilege to have known and worked with Luiz. He was not only an amazing engineer, but equally warm and kind-hearted. Sad. 😢
wired.com/story/google-mourn…
2
13
4,134
Marcel van Lohuizen retweeted
8 Feb 2023
#CueLang makes it very easy and elegant to add #monitoring to your system (in this example exposing @PrometheusIO scrape endpoints) across environments and deployments, with templating, validation, defaults and even enforcement.
Good talk by @cue_lang creator @mpvl_ @cfgmgmtcamp
1
4
23
2,217
9 Feb 2023
Great read!
“Transparent Telemetry for Open-Source Projects”
research.swtch.com/telemetry…
7
1,308
22 Sep 2022
A must read indeed: testscript is awesome! We use it everywhere in cue to quickly roll out comprehensive test sets that are also reusable for implementations in other languages as a bonus.
21 Sep 2022
This is a must read! #golang
1
3
8
15 Aug 2022
Welcome! Hugely excited to work with you all!
Announcement 🔈: the CUE project would like to officially welcome @carmatrocity, @rogpeppe, @mvdan_, and @ddm_tweets. Thread 👇
2
9
Marcel van Lohuizen retweeted
13 Mar 2022
10 years of my technical writing being critiqued by the Go team and I can still hear Rob et al telling me: "Drop Useful. Useful is subjective. And if it's not useful why would it be here anyway?"
java.util.Locale: ...
21
28
522
5 Feb 2022
Thanks @toshywoshy @KrisBuytaert Hoping to go back to IRL for next year!
5 Feb 2022
@KrisBuytaert @toshywoshy thanks very much for hosting us today, and for all the support and prep work that went in before. Very much appreciated!
1
1
3
Marcel van Lohuizen retweeted
5 Feb 2022
This is bad...
Malicious Kubernetes Helm Charts can be used to steal sensitive information from Argo CD deployments apiiro.com/blog/malicious-ku…
7
4
Reminder! Join @mpvl_ and @_myitcv tomorrow for "A practical guide to CUE: patterns for everyday use." Saturday 5 Feb at 1000 CET (UTC 1), part of @fosdem 2022!
x.com/cue_lang/status/148715…
On Saturday 5 Feb at 1000 CET (UTC 1), @mpvl_ and @_myitcv will be giving a talk in the Infra Management Dev Room at FOSDEM: “A practical guide to CUE: patterns for everyday use”
fosdem.org/2022/schedule/eve…
Look forward to seeing you there!
4
14
On Saturday 5 Feb at 1000 CET (UTC 1), @mpvl_ and @_myitcv will be giving a talk in the Infra Management Dev Room at FOSDEM: “A practical guide to CUE: patterns for everyday use”
fosdem.org/2022/schedule/eve…
Look forward to seeing you there!
11
27
Marcel van Lohuizen retweeted
26 Nov 2021
Off the back of this I’ve been playing with #cuelang today and my mind has been slightly blown apart by the idea of basing config generation on a lattice structure and treating types & values as the same thing.
Lots of great responses here on ways to test Helm values too 😄
26 Nov 2021
How are people unit testing Helm template files? This is getting sufficiently complex that I want to be able to assert that given certain values passed in, certain output comes out after templating (e.g. setting a value adds a flag to a deployment etc)
2
3
22