Architecture
Our brain-inspired architecture HRM powers deep, efficient reasoning through structured computation in latent space.
What is Latent Space?
Latent space is an internal representational space where abstract structure, relationships, and plans can form before they are expressed as words.
It is the internal workspace for thought—richer and more flexible than language.
What is Latent Space?
Latent space is an internal representational space where abstract structure, relationships, and plans can form before they are expressed as words.
It is the internal workspace for thought—richer and more flexible than language.
Why is Matters
Many problems require abstraction, planning, and internal structure that do not naturally fit into a token-by-token language process.
Reasoning in latent space enables deeper thinking, more efficient computation, and greater flexibility.
Input(Prompt)
Output(Language)
Reasoning is serialized through language.
The Limits of LLMs
Reasoning and language are mixed.
Conventional LLMs often reason through language space.
Even abstract problems must be processed step by step through token representations. This can make reasoning longer, more fragile, and more resource-intensive.
Input(Prompt)
Output(Language)
The HRM Approach
Reasoning is separated from language.
HRM performs structured computation in latent space first.
When communication is needed, the resulting thoughts are mapped into natural language.
Responsible for abstract, deliberate reasoning
Responsible for detailed computations
