Otters++: Energy-Efficient Optical Spiking Transformer Using Natural Device Decay
Researchers have developed Otters++, a spiking neural network architecture for Transformer models that exploits the natural signal decay of optoelectronic devices to perform time-to-first-spike computation without costly digital overhead. The system uses a custom Indium Oxide (In₂O₃) optoelectronic synapse whose physical decay behavior directly implements the temporal term normally requiring explicit calculation, and a hybrid training method bridges discrete spike events with standard gradient-based optimization. On the GLUE natural language benchmark, Otters++ achieves an average score of 84.17% while maintaining an energy advantage over prior spiking Transformer baselines.
Otters++ is a new optoelectronic spiking neural network (SNN) architecture designed to make Transformer models more energy-efficient by leveraging time-to-first-spike (TTFS) coding, where each neuron fires at most once per inference. A key innovation is repurposing the natural signal decay of a custom In₂O₃ optoelectronic synapse — typically considered a hardware imperfection — as the primary computational mechanism for the TTFS temporal decay term, eliminating the need for explicit digital computation of that term. To enable training at scale, the authors establish a layer-wise functional equivalence between Otters++ and a quantized neural network (QNN), allowing standard straight-through gradient estimation in the backward pass while preserving device-faithful SNN behavior in the forward pass, supplemented by model distillation. The training pipeline also incorporates measured device noise by sampling run-to-run variation, and the energy model is refined to account for device sharing and multi-hop communication overhead. Evaluated on the GLUE benchmark suite, Otters++ achieves an average score of 84.17%, improving over prior spiking Transformer baselines while retaining a meaningful energy efficiency advantage. The work demonstrates that physically grounded, hardware-aware TTFS computing can be made trainable and robust under realistic device conditions.
What's missing
The study relies on a custom In₂O₃ synapse device whose fabrication reproducibility, yield, and scalability to large model sizes are not fully characterized. Additionally, evaluations are limited to the GLUE benchmark; generalization to other tasks or modalities remains an open question.
What different sources said
- arXiv cs.AICenter
Otters++: A Time-to-first-spike Based Energy Efficient Optical Spiking Transformer
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.