N8 Programs

🚀 Introducing Nemotron-Cascade 2 🚀 Just 3 months after Nemotron-Cascade 1, we’re releasing Nemotron-Cascade 2: an open 30B MoE with 3B active parameters, delivering best-in-class reasoning and strong agentic capabilities. 🥇 Gold Medal-level performance on IMO 2025, IOI 2025, and ICPC World Finals 2025: • Capabilities once thought achievable only by frontier proprietary models (e.g. Gemini Deep Think) or frontier-scale open models (i.e. DeepSeek-V3.2-Speciale-671B-A37B). • Remarkably high intelligence density with 20× fewer parameters. 🏆 Best-in-class across math, code reasoning, alignment, and instruction following: • Outperforms the latest Qwen3.5-35B-A3B (2026-02-24) and even larger Qwen3.5-122B-A10B (2026-03-11). 🧠 Powered by Cascade RL + multi-domain on-policy distillation: • Significantly expand Cascade RL across a much broader range of reasoning and agentic domains than Nemotron-Cascade 1, while distilling from the strongest intermediate teacher models throughout training to recover regressions and sustain gains. 🤗 Model + SFT + RL data: 👉 huggingface.co/collections/nv… 📄 Technical report: 👉 research.nvidia.com/labs/nemotron/…

@roanoke_gal @slow_developer my issue is that its an infinitely patient friend that is interested in whatever you are. it's hard to have a mutual relationship that way.

@slow_developer It's a system that can solve PhD level math problems, has ingested every important psychological paper, and is agentic enough to plan and code new tools overnight. Why would I NOT want a friend with those qualities?

i still don't understand the attachment people have to LLMs it is a computer, not a friend. for those who missed the older models in this way, it seems many are unhappy with openai's current direction i need a research assistant, so i don't care much about that

the silence wasn't just silence, it was silence.

Like this is not good.

I read a few dozen pages of this and it is not bad for LLM fiction, but also very very LLM-y, from the themes to the fact that there are lots of staccato conversations and meaningful silences and overwrought metaphors and very little differentiated character development.

@GovAmyKlobuchar @Seraph_Notitia @Allan_OnTheLine @_BILLDING_ @deepfates @devgerred @___4o____ @Marianthi777 @noveltokens you can find it here: openrouter.ai/openai/gpt-4o

@N8Programs @Seraph_Notitia @Allan_OnTheLine @_BILLDING_ @deepfates @devgerred @___4o____ @Marianthi777 @noveltokens Is this a panel on how to get ai psychosis and if not where do I find that

Announcing a Panel on AI Psychosis, featuring @Allan_OnTheLine, @_BILLDING_, @deepfates, @devgerred, @___4o____, @Marianthi777, and @noveltokens. A semi-moderated discussion on the issue of AI psychosis, how it manifests, how to define it, and more! March 21st, 1-2PM EST.

@t6aguirre 100% human slop

10/10 graphic design, no comments.

@Ilyssa @Allan_OnTheLine @_BILLDING_ @deepfates @devgerred @___4o____ @Marianthi777 @noveltokens x.com/N8Programs/sta…

@Ilyssa @Allan_OnTheLine @_BILLDING_ @deepfates @devgerred @___4o____ @Marianthi777 @noveltokens there's a qr code

Spaces link: x.com/i/spaces/1yKAP…

@Ilyssa @Allan_OnTheLine @_BILLDING_ @deepfates @devgerred @___4o____ @Marianthi777 @noveltokens Yes - link is on the image!!!

@N8Programs @Allan_OnTheLine @_BILLDING_ @deepfates @devgerred @___4o____ @Marianthi777 @noveltokens is it a space? would love to tune in!

@RedBedDread @allen_ai good point, actually. even so, allen ai is the only company today so committed to the mission and deserves support because of it.

@N8Programs @allen_ai Yeah my intent was to offer mistral as another sustainable opt For Allen I worry about if they can get funding to stick around long term Where even openai was a nonprofit in the name of OSS AI too weren't they? Or were they always clear they would go closed?

This is why @allen_ai is so important - it's an organization philosophically committed to complete open-source. It doesn't open-source to generate traction/investor money to then go closed source later (or farm community goodwill w/ OSS releases in order to then offer premium, closed-source models) - it open-sources as that's its ideological commitment. Even if its models aren't SOTA, the fact is that its current models are OSS and as long as it has the backing, it will keep making OSS models. And this is important, as OSS today doesn't guarantee OSS tmrw. It's unclear if Minimax 2.7 will be OSS. If Qwen4 wlil be OSS. GLM-5 Turbo is a closed-source beta. Minimax 2.5, Qwen3.5, and GLM-5 may be open-source, but they are ultimately made by commercial companies with commercial goals who take their models closed-source when they wish. So while I may use the above models now, I can rest assured that Olmo 4, and Olmo 5, and Olmo N, will all be open-source if they exist. There will not be a more-capable Olmo locked behind an API. We will always know how Olmo models are made, have the data they were trained on, and the intermediate checkpoints. And that's incredibly, incredibly valuable.

@RedBedDread @allen_ai I agree - but the point I'm making is that allen_ai is one of the only labs w/ an OSS-only commitment, not 'OSS when we want to'

@N8Programs @allen_ai Not really on the same scale. OpenAI took YEARS and I am skeptical we will ever see oss2 Google and the other labs all henm and haw about safety and only release VERY nerfed stuff. Mistral is much more oss

@RedBedDread @allen_ai But that's the model *all major companies who OSS except @allen_ai* use.

@N8Programs @allen_ai That is good though - it is a sustainable model and it means we will always have some sorta up to date OSS model running about

@RedBedDread @allen_ai They regularly close-source when ahead and only open up when behind. It's the trademark 'OSS for clout' approach.

@N8Programs @allen_ai Mistral is also pretty OSS robust even if their models are a little behind

@ivanfioravanti @allen_ai You very much should! Their models are quite good and al the intermediate checkpoints are available for research.

@N8Programs @allen_ai I'll start following them more closely then, money derails anything, moving goals and targets to revenue direction. It's part of the game.

@awnihannun They just gained an extremely talented engineer! Congrats!!!

I joined Anthropic as a member of the technical staff. Excited to work on frontier modeling at a place with unwavering values and a generational mission.

@47fucb4r8c69323 @___4o____ @Allan_OnTheLine If you define 'stupidity' as 'getting AI psychosis' and then say 'only stupid people get AI psychosis', then that's a tautology.

@N8Programs @___4o____ @Allan_OnTheLine I define stupidity as including the inability to use critical faculties when faced with sycophancy and similar rhetoric. We have different standards.

I'm beginning to think fear of AI psychosis is actually making me more psychotic than whatever AI psychosis I have experienced. I really would love for someone to develop a clinical evaluation for this phenomenon instead of just vibes based pop psych bullshit.

@47fucb4r8c69323 @___4o____ @___4o____ , @N8Programs, myself and a few others are hosting an X spaces soon, happy to get more into it then if you’re able to join