Tensorized Engram: New Method for Efficient N-Gram Embeddings in Language Models
Two independent research papers propose methods to improve language models by explicitly incorporating n-gram (multi-token pattern) memory modules. Tensorized Engram (TN-gram) uses a compact tensor factorization approach for general language models, while MolGram applies similar principles to molecular language models for chemistry tasks. These approaches address a fundamental limitation where language models must implicitly learn recurring multi-token patterns, offering more parameter-efficient alternatives.
Two recent arXiv preprints present complementary advances in incorporating explicit n-gram memory into Transformer-based language models. TN-gram proposes a tensorized memory module using Canonical Polyadic decomposition to represent n-gram embeddings with shared factors, enabling different n-gram orders to share underlying latent structures while reducing parameters compared to prior Engram-style approaches. Separately, MolGram addresses the specific challenge of molecular language models processing SMILES strings, where character-level tokenization fragments chemically meaningful motifs; it integrates a conditional n-gram memory module with scalable hash lookups to inject local pattern context into hidden states. Both papers demonstrate that explicit local pattern memory serves as an efficient inductive bias, with MolGram showing performance improvements while using 3× fewer parameters than baseline models. The parallel development of these techniques across general and domain-specific language modeling suggests growing recognition that standard token-level embeddings inadequately capture recurring multi-token structures.
What different sources said
- arXiv cs.LGCenter
Tensorizing Engram: Sharing Latents Across N-Gram Embeddings is Beneficial in LLMs
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.