CEO of Sapient Intelligence. Building leaner architectures for deeper intelligence.
Singapore
Joined July 2025
- Tweets38
- Following42
- Followers5.6K
- Likes128
Pinned Tweet
The HRM-Text paper is now available 🎉
HRM-Text explores a different approach to language model pretraining: hierarchical recurrent computation, task-completion training, and latent-space reasoning.
At just 1B parameters, HRM-Text achieves competitive performance with dramatically lower training cost and data requirements.
1B parameters
40B unique tokens
~1 day of pretraining
~$1000 training cost
Another step forward in continual learning and adaptation🚀
We’re happy to share that two papers from Sapient Intelligence have been accepted to NeurIPS 2026!
The accepted work contributes to a broader research agenda around optimization, adaptation, and efficient learning in complex systems, all of which are key to building increasingly capable and autonomous learning systems.
See you at NeurIPS 🎉
Guan Wang retweeted
Introducing PRAXIST Beta, your autonomous research team🚀
Define the objective, constraints, and what success looks like. PRAXIST discovers the path.
From there, PRAXIST takes on the experimental research loop. Multiple Research Peers explore competing approaches in parallel, share useful findings, and build on accumulated evidence across experiments and generations.
In partner-provided environments, PRAXIST reached a 100% safe-landing rate in a rocket simulation and reduced accumulated error in an industrial SLAM system from 9.37 centimeters to 5.01 centimeters.
From robotics control to quantitative finance, PRAXIST has delivered measurable advances across fundamentally different problems by combining a shared core research architecture with domain-specific tools, knowledge, constraints, and evaluators.
More approaches explored. Faster iteration. Greater R&D capacity without proportionally increasing specialist headcount.
Really enjoyed our conversation about what’s next for HRM. Truly appreciate the continued support from @bendee983 and the @VentureBeat team!
Last year, @Sapient_Int introduced the Hierarchical Reasoning Model (HRM) as a proof-of-concept architecture that could replace the inefficient token-by-token mechanism of Transformers in reasoning tasks. This year, they moved toward more practical applications with HRM-Text, a newer version of the model that can perform text reasoning tasks.
The main change HRM brings is that it replaces next-token prediction with a recursive hierarchical mechanism that reasons in latent space. It is faster and much more efficient.
Key highlight:
A 1B-param HRM-Text model, trained on 40 billion tokens and $1,500 worth of compute, outperformed much larger models trained on trillions of tokens on key reasoning benchmarks. It is also much more efficient because it doesn't
The main takeaway is that autoregressive, next-token-prediction training is very inefficient and forces the model to memorize trillions of tokens' worth of reasoning traces to learn general reasoning. HRM reasons in latent space and isn't bound by NTP.
I spoke to @makingAGI about the promises of latent-space reasoning and the next steps for HRM and HRM-Text.
Guan Wang retweeted
Excited to see our community building with HRM-Text! Can’t wait to see where you take it next.🚀
Whether you’re reproducing the benchmarks, testing it out in your own field, or building something entirely new, we’d love to hear about it.
Drop your ideas, questions, and experiments below, and tag @Sapient_Int so we can follow along and cheer you on!🎉
Huge congrats on reproducing the HRM-Text results! 🚀Pretraining an LLM on just 16 GPUs is a massive win. 👏#AI #NeuroAI #LLM #HRMText
Reproduced HRM-Text XL (1B).
Training completed in ~38 hours wall-clock on 16 H200 GPUs, and evaluation performance matches the numbers reported in the paper.
Great job, team!
W&B report:
api.wandb.ai/links/MDEQGA/70…
Congratulations to the team! 👏Excited to see new explorations building on the HRM architecture and pushing recursive reasoning further.
🧠We introduce "Generative Recursive Reasoning"!
Recursive Reasoning Models like HRM, TRM, and Looped Transformers are deterministic — same input, same reasoning, every time. They collapse the entire space of plausible reasoning paths into a single attractor.
Our model GRAM (Generative Recursive reAsoning Models) turns recursion itself into a stochastic latent trajectory. Multiple hypotheses, alternative solution strategies, and inference-time scaling not just by depth, but by width — parallel trajectory sampling.
And here's the kicker: the same formulation that gives us conditional reasoning p(y|x) also makes GRAM a general generative model p(x).
With only 10M params:
• Sudoku-Extreme: 97.0% (TRM 87.4%)
• ARC-AGI-1: 52.0%
• ARC-AGI-2: 11.1%
• N-Queens coverage: 90%+
📄 Paper: arxiv.org/abs/2605.19376
🌐 Project page: ahn-ml.github.io/gram-websit…
w/
Junyeob Baek @JunyeobB (KAIST),
Mingyu Jo @pyross0000 (KAIST),
Minsu Kim @minsuuukim (KAIST & Mila),
Mengye Ren @mengyer (NYU),
Yoshua Bengio @Yoshua_Bengio (Mila),
Sungjin Ahn @SungjinAhn_ (KAIST)
Paper: arxiv.org/abs/2605.20613
GitHub: github.com/sapientinc/HRM-Te…
Hugging Face: huggingface.co/sapientinc/HR…
Guan Wang retweeted
In this benchmark deep-dive, Sapient’s founders William and Guan are joined by research team members Changling and Yasin to unpack HRM-Text’s performance across MATH, DROP, ARC-Challenge, and MMLU. 📊
Beyond the scores, they discuss what each benchmark measures, how HRM-Text compares with larger models, and why efficiency matters.
Watch the full discussion to learn more about HRM-Text and Sapient’s leaner path toward general intelligence.
Guan Wang retweeted
HRM-Text 101 is here.
This tutorial takes you from zero to one: from setup to fine-tuning to evaluation.
Download the base checkpoint.
Fine-tune it on a real task.
Evaluate the results.
End to end, on a single GPU.
Watch the tutorial and start building with HRM-Text.
Guan Wang retweeted
Introducing HRM-Text.
An ultra-lean 1B-parameter reasoning language model designed to deliver strong general performance with a fraction of the data, compute, and infrastructure.
Trained on just 40B structured tokens, HRM-Text achieves competitive performance while using ~1/1000 of the training data of comparable models.
The kicker? The full model trains in roughly one day on a $1,000 budget.
This opens the door to a new generation of AI that is powerful, accessible, and radically easier to adapt. Theories and research concepts once deemed too expensive to test are officially back in the game.
Sapient Intelligence invites you to help us shape a new paradigm for general intelligence.
Guan Wang retweeted
Tomorrow, we will unveil a new path to general intelligence.
Lean. Powerful. Efficient.
The countdown is on⏳
Guan Wang retweeted
Behind the code, there is a specific kind of expertise.
We are a team of researchers and engineers rooted in the labs of Tsinghua University, University of Cambridge, University of Alberta, Carnegie Mellon University, and Peking University—with experience at DeepMind, DeepSeek, xAI, and more.
We've seen the limits of the current AI architectures firsthand from within the organizations that scaled them. Now, across three countries, we are building an alternative.
We aren't just shipping another wrapper; we are shipping a new fundamental architecture.
Guan Wang retweeted
🔥It’s official-Sapient HRM Discord Community is now live！
This is a place to discuss, connect, and collaborate as we shape HRM’s future together. We will be sharing our latest work, releases, and tips, as well as hosting Q&A sessions💬💬
Hop on this journey with us as we push the boundaries of what HRM and AGI at large can achieve！🙌
➡️Join us on Discord here discord.gg/sapient
Thanks to @arcprize for reproducing and verifying the results!
ARC-AGI-1: public 41% pass@2 - semi private 32% pass@2
ARC-AGI-2: public 4% pass@2 - semi private 2% pass@2
Due to differences in testing environments, a certain amount of variance in results is acceptable. According to tests run on our infrastructure, the open-source version of HRM on our GitHub can achieve a score of 5.4% pass@2 on the ARC-AGI-2. We welcome everyone to run it on your own infra and share your scores~
This is our first submission to the leaderboard, and it's a good starting point. We appreciate everyone for your support and feedback on HRM, both before and after our appearance on the ARC leaderboard. All of this encourages and motivates us to improve.
The hierarchical architecture is designed to resolve premature convergence in long-horizon tasks, like master-level Sudoku that takes hours for humans to solve. See the comparison with a simple recurrent Transformer. Such a long chain might not be essential for ARC problems, and we only used a high-low ratio of 1/2. Larger ratios are often needed for optimal performance for Sudoku problems.
In the case of ARC-AGI, the success of HRM is a testament to the model's ability to exhibit fluid intelligence - that is, its capability to infer and apply abstract rules from independent and flat examples. We are glad it was discovered in a recent blog post that the outer loop and data augmentation are essential for this ability, and we especially thank @fchollet @GregKamradt @k_schuerholt for pointing this out.
Finally, we are accelerating the iteration of the HRM model and continuously pushing its limits, with good progress so far. At the same time, we believe the hierarchical architecture is highly effective in many scenarios. Moving forward, we will make further targeted updates to the architecture and validate it on more applications. We will also release an FAQ to address the key questions raised by the community. 🧠 Stay tuned!
❤️ Thanks for highlighting our HRM paper! Apologies for any confusion—we're working on clarifying this thoroughly. Stay tuned for updates!
Thanks for featuring us!😃
My story on HRM with comments from @makingAGI, CEO of Sapient Intelligence
xitter.cf/VentureBeat/status/194…