Doc-to-Atom: New Framework Decomposes Documents into Semantic Knowledge Units for Efficient LLM Processing
Researchers have proposed Doc-to-Atom (Doc2Atom), a compositional memory framework that decomposes documents into semantically typed 'knowledge atoms,' each compiled into a small independent adapter for large language models. The work builds on prior context distillation approaches like Doc-to-LoRA, which compress document information into model parameters but suffer from interference and poor scalability when a single adapter must serve all queries. Doc2Atom addresses these limitations by routing only relevant atoms to a query-specific adapter, outperforming baselines on six QA benchmarks while reducing memory costs.
Doc-to-Atom (Doc2Atom) is a new parametric memory framework introduced to tackle the computational challenges of long-document understanding in large language models (LLMs). Standard attention mechanisms scale quadratically with sequence length, making inference on long documents both slow and memory-intensive. Prior work on context distillation, including Doc-to-LoRA, attempted to address this by compressing document context into a single LoRA adapter per document in one forward pass, but this monolithic approach causes irrelevant-query interference and limits compositional recall. Doc2Atom instead decomposes each document into semantically typed knowledge atoms, compiling each into an independent micro-LoRA adapter paired with a provenance retrieval key. At inference time, a lightweight query router selects and assembles only the atoms relevant to a given query, injecting the resulting adapter into a frozen base model. The entire pipeline is trained end-to-end via a multi-objective distillation framework. Evaluated across six diverse question-answering benchmarks, Doc2Atom outperforms Doc-to-LoRA baselines while also reducing the memory footprint of document internalization.
What's missing
The paper does not report results on documents beyond a specified length range, leaving open questions about scalability to extremely long corpora or multi-document settings. The computational overhead of the query router itself and the sensitivity of performance to the granularity of atom decomposition are not fully characterized. It is also unclear how the system performs on non-English or domain-specific documents outside the tested benchmarks.
What different sources said
- arXiv cs.CLCenter
Doc-to-Atom: Learning to Compile and Compose Memory Atoms
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.