
Founding Member of Technical Staff @RicursiveAI | AI × Chip Design Prev: Director @Cadence, SWE @Google, Alphachip
San Francisco Bay Area
Joined April 2020
- Tweets54
- Following12
- Followers141
- Likes108
Jiwoo Pak retweeted
We just set the U.S. open-weight SOTA on agentic coding. Tomorrow, @DZhang50 goes behind the scenes on Ripple-65B Preview with @zainhas of @togethercompute: dense over MoE, ablations, and the RL runs.
📆 Nov 17 · 11am PT
👉 luma.com/together-1kx8
How do you build a small coding model from scratch?
@RicursiveAI 's Dan Zhang breaks down the architecture and decisions behind Ripple-65B Preview with @zainhas: dense over MoE, the ablations that held up (and didn’t), and what happened inside the RL runs.
Tuesday, Nov 17 . 11 A.M. PT / 2 P.M. ET
Jiwoo Pak retweeted
A fantastic week at @ICCAD 2026. The Ricursive booth was as busy as ever. It was great meeting so many talented researchers, engineers, and field experts.
We hire those who are the very best at what they do, and who do it with radical humility and kindness. Once you're here, you'll design algorithms, run large-scale experiments, and ship work that changes how chips get designed.
Join us: ricursive.com/careers
Our first hackathon was a huge success! Thanks to everyone who joined and built with us. 🙌
Last Friday we hosted the Ricursive Intelligence After Dark Hackathon, our first hackathon centered on our open-weight model release, Ripple-65B Preview. We brought together an incredible group of researchers, engineers, and builders to push the model in new directions. We saw some amazing projects spanning everything from CUA and Weight-Space Programming to model watermarking.
Huge thanks to everyone who came out, built something, and made the night memorable! This is just the beginning!
2
27
Jiwoo Pak retweeted
Last Friday we hosted the Ricursive Intelligence After Dark Hackathon, our first hackathon centered on our open-weight model release, Ripple-65B Preview. We brought together an incredible group of researchers, engineers, and builders to push the model in new directions. We saw some amazing projects spanning everything from CUA and Weight-Space Programming to model watermarking.
Huge thanks to everyone who came out, built something, and made the night memorable! This is just the beginning!
Jiwoo Pak retweeted
Last week, we released Ripple-65B Preview. A 65B dense LLM that set a new U.S. open-weight SOTA on agentic coding.
Join our very own @DZhang50, Founding Member of Technical Staff, and @togethercompute's @zainhas to learn how we did it.
📆 Tuesday, Nov 17 · 11am PT / 2pm ET
👉 Register at luma.com/together-1kx8
How do you build a small coding model from scratch?
@RicursiveAI 's Dan Zhang breaks down the architecture and decisions behind Ripple-65B Preview with @zainhas: dense over MoE, the ablations that held up (and didn’t), and what happened inside the RL runs.
Tuesday, Nov 17 . 11 A.M. PT / 2 P.M. ET
Jiwoo Pak retweeted
Ripple-65B Preview is now available on ollama!
ollama.com/library/ripple-pr…
Ripple 65-B Preview from @RicursiveAI is on Ollama. Built for agentic coding, with up to 1M tokens of context. Run it locally with full zero data retention.
Jiwoo Pak retweeted
Last week we released Ripple-65B Preview. This Friday, come build with it.
Ricursive After Dark Hackathon: Fri Nov 13, 3PM–10PM, at our Palo Alto office. Compute, dinner, and hardware prizes.
Applications close Thursday, November 12 at noon PT. Accepted participants will be notified by email Thursday night with the address and full details.
luma.com/qnsar3a2?utm_source…
Jiwoo Pak retweeted
the community reaction to our model has been amazing!
⚙️ TESTED LOCALLY: Ripple-65B Preview, a 65B open-weight model from @RicursiveAI, on my 64GB M5 Pro Mac mini
It's dense, Apache 2.0 and pretrained from scratch. I ran the community MLX 4-bit build.
- Steady 7 tok/s from 512 to 32K tokens
- Peaks at 40.5 GB, room to spare on 64GB
- 53 for 53 on my local tests, hidden facts out to 64K inc
- Runs about as fast as this chip allows for a dense 65B
The community had an MLX build up within a day.
Good day for owning your intelligence.
Jiwoo Pak retweeted
GMI Cloud is proud to be a Day 0 launch partner for Ripple-65B Preview from @RicursiveAI 🚀
It's a 65B dense model pretrained from scratch, and it sets a new state of the art for U.S. open-weight models on agentic coding tasks, ahead of much larger models.
Try it now 👇
Today we release Ripple-65B Preview, our 65B dense LLM we pretrained from scratch.
Ripple sets a new state-of-the-art for U.S. open-weight models in agentic coding tasks, surpassing significantly larger models including Muse Spark Open-816B, Beam-501B, Inkling-975B, and Nemotron 3 Ultra-550B.
Learn more: ricursive.com/ripple
Jiwoo Pak retweeted
Ripple was built on a $32 million compute budget, one-tenth the cost of the previous open-weight SOTA on agentic coding. We're grateful to the open-source community, whose tools and research helped make this possible.
You can access the model here:
HuggingFace: huggingface.co/ricursiveAI/R…
GMI Cloud: console.gmicloud.ai/model/ll…
OpenRouter: openrouter.ai/gmicloud/rippl…
You can also access Ripple through Together AI and Fireworks starting later today. Full tech report coming soon!
Jiwoo Pak retweeted
On the pretraining front, we achieved a 16x compute efficiency win on code relative to our Olmo 3 baseline through improvements in scaling laws, model architecture, and data.
Jiwoo Pak retweeted
Our in-house posttraining stack and synthetic data pipelines powered an efficient mixed RL climb in agentic performance. We reached 80.5 on TB2.1 with only 1.4M rollouts, roughly 70x fewer rollouts than Reflection Beam, the prior SOTA.
Jiwoo Pak retweeted
Ripple-65B Preview was trained on 12 trillion tokens, supports up to 1M context length, and achieves:
- 80.5% on TerminalBench 2.1
- 79.1% on SWE-Bench Pro v2.0-Hard
- 44.3% on DeepSWE v1.1
All while fitting locally on a MacBook or DGX Spark.
Jiwoo Pak retweeted
Today we release Ripple-65B Preview, our 65B dense LLM we pretrained from scratch.
Ripple sets a new state-of-the-art for U.S. open-weight models in agentic coding tasks, surpassing significantly larger models including Muse Spark Open-816B, Beam-501B, Inkling-975B, and Nemotron 3 Ultra-550B.
Learn more: ricursive.com/ripple
Jiwoo Pak retweeted
We're hiring! 🚀
Our team is growing, and we're looking for three incredible people to join us:
🗓️ An Executive Assistant to support our co-founders
🤝 Our first People Ops hire
✍ A Technical Writer who can double as a head of comms
We hire those who are the very best at their craft, with radical humility. If that's you, we want to meet you: ricursive.com/careers.
Jiwoo Pak retweeted
Great to see Ricursive's own @carolinetrippel give an invited talk at #ICLAD2026 🎙️.
Her research tackles hardware bottlenecks head on, using AI + Formal Methods to turn 1-hr timeouts into sub 15-sec checks. Proud to have Caroline guiding Ricursive's DV direction.
Jiwoo Pak retweeted
3 packed days at @DACconference!
The energy at Booth #1049 was unmatched. Huge thanks to everyone who stopped by to learn more about Ricursive, brainstorm on accelerating chip design, or just pick up some cool swag. Until next year!
Jiwoo Pak retweeted
Proud to join @Microsoft, @nvidia, @a16z, @PalantirTech, and 230+ tech leaders as a signatory supporting Open Weights and American AI Leadership.
An open, secure ecosystem is essential to accelerating hardware-software-model co-design and democratizing custom silicon.
Since launching last week, more than 230 companies and organizations from across the tech sector have signed the "Open Weights and American AI Leadership" open letter. We want to thank these partners for standing up and publicly supporting broader access to AI innovation. A special thanks to @nvidia, @a16z, and @PalantirTech for working with @Microsoft on this effort.
These signatories understand that America’s AI leadership will not depend on the success of our frontier models alone, but on our ability to build a strong, secure, and open ecosystem that diffuses AI into every sector. We look forward to continuing to work with our partners and with policymakers to build that open ecosystem in a way that benefits American businesses, empowers American workers, and strengthens the American economy.
Jiwoo Pak retweeted
"AI inference demand is expected to grow 10,000x over the next 5 years."
Yesterday at @DACconference, @annadgoldie took the stage to share how we break today's deadlock between model and hardware. That's what we're doing at Ricursive, using our very own model.
📍 Stop by Booth #1049 to say hi.
Jiwoo Pak retweeted
The event was scheduled to end at 8pm. The conversations ran well past it.
Ricursive's "Chips & Pints" outgrew the room we booked, so we took over the brewery. Thanks to everyone who battled the Long Beach heat after day 1 of @DACconference to spend Monday evening with us 🍻.





























