New Memory Architecture Enables Efficient Deep Recursive Binding in Neural Networks
Researchers have proposed Orthogonal Subspace Carving (OSC), a memory architecture that enables deep recursive symbolic binding within a fixed memory footprint. OSC bridges the gap between Tensor Product Representations, which offer structural fidelity but scale exponentially, and Vector Symbolic Architectures, which maintain constant dimensionality but lose fidelity through noisy compression. The work addresses a fundamental bottleneck in neuro-symbolic AI by decoupling memory cost from structural depth, potentially enabling more scalable symbolic reasoning in neural models.
A paper accepted to the 43rd International Conference on Machine Learning (ICML 2026) introduces Orthogonal Subspace Carving (OSC), a novel memory architecture designed to overcome the dimensionality limitations of existing approaches to recursive symbolic binding. Traditional Tensor Product Representations (TPRs) preserve structural fidelity needed for symbolic reasoning but suffer from exponential memory growth as recursive structures deepen. Vector Symbolic Architectures avoid this by maintaining constant dimensionality, but at the cost of accuracy due to superposition noise. OSC resolves this trade-off by projecting fillers onto the null space of a role basis before aggregating them into a fixed order-p tensor, enforcing geometric orthogonality between bound structures within a static memory trace. This design decouples tensor order from structural depth, allowing arbitrarily deep recursive binding without increasing memory size, and enables component vectors far smaller than the memory tensor itself. The authors also demonstrate that TPR is a special case of binding within Clifford algebra and provide a Clifford algebraic formulation of OSC, situating the work within a broader mathematical framework.
What different sources said
- arXiv cs.LGCenter
Recursive Binding on a Budget: Subspace Carving in Order-p Tensor Memories
Related
Gut Bacteria Enzyme Found to Break Down Heat-Processed Food Compounds, Producing Novel Biogenic Amines
Researchers have discovered that an enzyme in common gut bacteria can degrade N-epsilon-carboxymethyllysine (CML), a compound formed during thermal food processing, producing previously unknown biogenic amines. The enzyme, ornithine decarboxylase SpeC from enterobacteria, acts on CML and related modified lysine derivatives through a low-level 'underground' catalytic activity. This finding suggests a previously unrecognized communication axis between thermally processed dietary compounds and gut microbial physiology, with potential implications for host health.
Full-Length Gene Sequencing Reveals Two Distinct Bacterial Communities in Black-Legged Ticks Expanding Into Canada
Researchers used Oxford Nanopore full-length 16S rRNA gene sequencing to characterize the microbiome of Ixodes scapularis black-legged ticks collected in Nova Scotia, Canada, distinguishing between tick-adapted bacteria and environmentally acquired bacteria. The study comes as I. scapularis — the primary vector of Lyme disease — is rapidly expanding northward into Canada due to climate change. The findings suggest that environmentally derived bacteria in tick microbiomes are not mere contamination, which has implications for how tick microbiome data is collected and interpreted across surveillance studies.
Study Identifies Metabolic Link Between Cell Envelope Stress and Biofilm Formation in Bacteria
Researchers have discovered that the metabolite acetyl-CoA directly inhibits enzymes that degrade the bacterial signaling molecule c-di-GMP, connecting cell envelope biosynthesis stress to biofilm formation in Pseudomonas aeruginosa. The study found that sub-inhibitory concentrations of antibiotics targeting early peptidoglycan biosynthesis — but not other antibiotic classes — elevate c-di-GMP levels by reducing phosphodiesterase activity, with acetyl-CoA competing for the enzyme active site. Because the relevant enzyme domain is broadly conserved across bacterial species, this checkpoint mechanism may be widespread and could have implications for understanding antibiotic-induced biofilm responses.