We are a small, independent research group working on fundamental AI research. With enough compute to actually run the experiments.

Research priorities

• Main priority: Investigate reasoning LLMs: RLVR, GRPO, SDPO.
• Secondary priorities:
  • Investigate ARC v3.
  • Learn to pre-train and post-train LLMs, and develop the know-how and infrastructure for it.
  • Investigate fancy ML-based quantitative strategies.

Strategy

• Funding is secured for at least the next 4 years.
• Within that timeframe, do interesting research (LLM/reasoning/RLVR, and ARC v3).
• This should open new financing and recruiting opportunities (hopefully elevating the lab from GPU-poor to GPU-middle-class to GPU-rich).
• Move on to more challenging benchmarks.

Values

Intellectual curiosity, high intrinsic motivation (obsession), perseverance. Fast experimentation process.

We are driven primarily by the desire to reverse engineer intelligence and build machines according to those principles.

Regarding AI safety: we believe that after reaching a certain level of capability, AI safety will become a critical issue and is likely to constitute an existential threat. But first we want to carry out the capability research and reassess as we go. We will have a process where AI safety can be made the top priority.

Compute