Skip to main content
Search

Technical University of Denmark

10 related articles

Subscribe to updates
  1. bioRxiv

    REXzyme translates a chemical reaction into an enzyme sequence, and its formate dehydrogenase design C20 reaches 2.5 times wild-type activity

    Enzyme engineering has relied on directed evolution, which screens thousands of variants even though fewer than 5% of single mutations are beneficial. REXzyme recasts enzyme design as translation: an encoder reads a reaction written in SMILES and a decoder generates an amino acid sequence token by token, after training on about 21 million reaction-enzyme pairs. Steered toward the reverse reaction of formate dehydrogenase, the reduction of CO2 to formate, all 20 designs generated after family-specific fine-tuning and reinforcement-learning alignment expressed as soluble proteins and showed activity, and design C20 reached a specific activity of 80.8 mU/mg, 2.5 times that of wild-type CtFDH.

  2. Nanoparticle safety stops being a material-only question: a host-response axis lifts biosafety AUC from 0.66 to 0.90

    Nanoparticle neurological safety is hard to predict from material properties alone, because exposure conditions, disease context and host responses reshape outcomes. Across 893 experimental contexts from 599 neurological studies, pro-inflammatory, anti-inflammatory/resolution and apoptosis-associated responses were harmonized into a Conserved Response Axis and added layer by layer to material descriptors. With a Random Forest model, biosafety ROC-AUC rose from 0.66 with material properties alone to 0.88 with exposure and biological context and 0.90 with cellular responses; recovery rose from 0.64 to 0.77 and 0.78.

  3. Swap one Schelling agent for an LLM and the segregation trajectory holds, but runtime climbs from 2 seconds to 8,917

    In a Mesa implementation of the Schelling segregation model, one agent's symbolic rule for classifying neighbors is replaced by a locally served Qwen3.5 model that classifies through tool calls, while the rest of the dynamics stays fixed. Statistical model checking with MultiVeStA shows the expected ratio of happy agents rising from about 0.65 to about 1.0 within 60 steps in both variants, with point-wise t-tests finding no difference, yet the LLM version takes 8,917 seconds against about 2 seconds. The 0.8b and 2b models show semantic errors, wrong tool-call counts or stuck generation in unit-style checks, while 4b and 9b pass all of them.

  4. arXiv

    Epidemic parameters without training a network: one least-squares solve, errors down up to four orders of magnitude

    Parameter estimation for epidemic compartment models no longer needs neural-network training. Physics-informed regression previously estimated time derivatives from data, and that derivative estimate became the accuracy bottleneck when sampling was coarse or noise large, while point estimates carried no uncertainty. The new physics-informed weak regression integrates the model in time instead, keeps the equations linear in the parameters, estimates the unknown initial state jointly, and reaches fourth-order accuracy with cubic-spline quadrature; across the eleven DINN benchmarks it is the most accurate estimator in 15 of 22 model/noise settings, one to four orders of magnitude more accurate than the reported DINN results where sampling resolves the dynamics, at about a millisecond per fit.

  5. arXiv

    Ship track forecasts move to Transformers: 0.18 km average error, prediction interval in one pass

    Ship trajectory forecasting no longer needs recurrent networks or repeated sampling. Earlier AIS-based models either used only the target vessel's history or relied on clustering other vessels, and approximated probability intervals through multiple forward passes. Here the vessel's past positions and a local KDE heat map are fed to a Transformer trained with quantile loss, reaching about 0.18 km average displacement error across three Danish water areas while emitting a prediction interval in a single forward pass.

  6. arXiv

    RAGenome scales a retrieval-based genomic language model to 13,312 nucleotides, lifting gene-finding MCC from 0.45 to 0.60

    The authors introduce RAGenome, the first retrieval-based genomic language model: it retrieves homologous sequences from a 100-vertebrate whole-genome alignment, discards the 73% of tokens that are gaps, and scales pretraining context to 13,312 nucleotides, 100 times longer than existing MSA-based gLMs; it raises gene-finding MCC from GPN-Star's 0.45 to 0.60 and reaches 0.87 AUROC on pathogenic-variant prioritization, unifying long-range and evolutionary signals with 168M parameters and 16,032 GPU hours.

  7. arXiv

    Polytopal Neural Networks constrain latent representations to learned polytopes, retaining 96.9% of unconstrained accuracy on ImageNet-1k

    The authors propose Polytopal Neural Networks (PNNs), which constrain each layer's latent representation during training to a convex combination of learned archetypes so that the explanatory coordinates are the same coordinates the next layer receives; across image classification and reconstruction benchmarks, PNNs approach unconstrained networks with small accuracy loss, outperform VQ-VAE and Dirichlet VAE in unsupervised reconstruction, and offer a VQ training route without EMA updates, commitment loss, or straight-through estimator.

  8. arXiv

    NSGA-II optimization of resource-level handover policies cuts cost by 37% and waiting time by 58% on average across synthetic and real event logs

    The work introduces the first approach for multi-objective optimization of resource-level handover policies in business processes: taking a Multi-Agent System (MAS) simulation model as input, it uses NSGA-II to search for Pareto-optimal person-specific handover policies, and across five synthetic Loan Application variants and seven real event logs it reduces cost by an average of 37% and waiting time by 58% relative to the as-is process, outperforming availability, random, lowest-cost (LC), and shortest-processing-time (SPT) heuristic baselines in most settings.

  9. arXiv

    XACT learns sparse attribution masks over invertible time-frequency transform coefficients, highlighting fewer spurious features than baselines on synthetic data and yielding sparse, structured explanations on two real-world datasets

    The authors propose XACT, a general framework that learns sparse attribution masks over coefficients from arbitrary invertible time-frequency transforms (STFT, continuous wavelet transform, discrete wavelet transform), and extend the virtual inspection layer from the STFT to both wavelet transforms so that LRP can generate explanations in these representations; on a synthetic dataset XACT produces precise explanations and is less prone to highlighting spurious features than the tested baselines, and across two real-world datasets it produces sparse and structured explanations, although no method performs best across all quantitative evaluation criteria.

  10. arXiv

    Does per-frame early exit pay off? A dynamic-depth speech enhancer lands on the same latency-quality frontier as static models on STM32N6, with a 26 microsecond per-frame policy

    The work supervises every intermediate depth of one causal speech enhancement model and fine-tunes its output heads so deeper outputs are never worse than shallower ones, yielding a family of static models that are more Pareto-efficient than equivalently-sized counterparts trained from scratch on the same budget (up to 0.11 higher PESQ at equivalent compute, matching the best PESQ at 30% less compute); after int8 quantization on an STM32N6 microcontroller, the dynamic enhancer lies on the same latency-quality frontier as the static models, with the policy running in 26 microseconds per frame on the companion Cortex-M55 and splitting the enhancer into separate NPU graphs adding 2.2% latency overhead.

You’re all caught up