A collection of papers exploring how knowledge, reasoning and systems evolve and how they can be rethought for an AI-first world.
We don't just build on AI research — we contribute to it. Our work argues for a new architecture for how machines represent and reason with knowledge.
The industry is racing to make models bigger and faster. We think the harder and more important problem is how knowledge is represented, recalled and used in the first place.
Most AI today predicts the next most probable token in a sequence — extraordinary at it, and so fluent the output feels like understanding. But fluency isn’t understanding. Models work brilliantly at the token level and retrieval works over whole documents, yet neither aligns with the level at which knowledge actually lives: the concept.
A single document holds many concepts; a single concept is scattered across many documents. That mismatch is architectural — not something more parameters or a bigger context window will patch. It’s a research problem, and one that’s been developing in academic work for over a decade. That’s the problem we’ve taken on.
Two published papers, one argument: intelligence lives at the level of concepts — how they’re held across an ecology, and how they compress and combine.
Designing AI for Internal Memory, Proximal Ecosystems, and External Unknowns
Human knowledge lives in three domains: Internal — what you hold yourself; Proximal — the accessible resources around you; and Frontier — the external unknowns. Creativity arises from optimal synthesis between the three — the surprise gradient — and an AI that collapses them into one retrieval problem fails at all of them. This is the frame everything we build sits inside.
The formal core of the programme.
Intelligence operates via atomic units that function as compression layers, dynamically composed into novel configurations. The Compression Calculus formalises representational efficiency across ten domains; the Compounding Cascade multiplies those ratios across abstraction layers. Composition of atoms remains unsolved (F1 0.13) — named openly as the next problem.
Special thanks to Benjamin Brey, Dr Sharon Jheeta and Priyanka Kocchar.
Concept-level representation cut message length by 46.2% while preserving meaning, and atomic retrieval hit 100% Recall@5 against 91% for chunk retrieval, using 47.5% less retrieved context. Composition — reliably assembling atoms into more complex concepts — is the piece we haven’t solved: F1 0.13, named openly as the next problem.
We share our thinking with the research community because a new architecture for intelligence is bigger than any one company. We publish our frameworks, our findings, and the problems we haven't solved yet.
Our approach sits within a growing body of research — neurosymbolic AI, cognitive science and knowledge representation all point towards structure and meaning, not scale alone.
The idea that intelligence depends on reusable, composable units isn’t ours alone — it runs through decades of research. Cognitive science shows that experts think in compressed chunks, not raw detail. Information theory ties compression to prediction and understanding.
Neurosymbolic AI works to combine the fluency of neural models with the structure and traceability of symbolic reasoning. And recent work on compositional generalisation and library learning shows systems discovering and reusing their own abstractions.
Our contribution is to bring these threads together into one measurable framework — and to build on it. We’re advancing a frontier alongside a scientific community increasingly converging on the same insight.
Type 2.5 today. The next step is the one worth arguing about.
In 2020, Henry Kautz set out a taxonomy for how neural and symbolic systems can be combined — a ladder running from systems that merely bolt the two together, through those where each calls the other, to systems where symbolic reasoning is genuinely embedded in the neural substrate.
Most of what ships today sits near the bottom of that ladder: a language model with a retrieval step attached. SeKondBrain sits at Type 2.5 — the concept graph is not a lookup the model calls, it is structure the reasoning runs through. Meaning is held symbolically and stably; reasoning is traceable; the neural layer reads the world and the symbolic layer holds what it means.
We are not at the top of the ladder, and we will not claim to be. The step above is where symbolic structure and neural computation stop being two systems in conversation and become one. That is the problem we are working on — and the next section is where we are with it.
The concept graph isn’t a lookup the model calls — it’s structure the reasoning runs through, held symbolically and traceably. We’re not at the top of Kautz’s ladder, and we won’t claim to be.
Three active workstreams. One has data; none has a victory lap.
Our published work showed retrieval over concepts is sharper and lighter than retrieval over documents. What it didn’t solve is composition — reliably assembling atomic concepts into more complex ones. Concept-Grounded Attention is our line of attack, grounding attention itself in the concept graph.
Read the work, or follow along as the next results land. This is the science behind everything we build — read the papers in full, see how the thesis becomes a product on the Platform, or follow the research as it grows with the Community.
