Explore the vast range of titles available.

Building a large-scale quantum computer requires effective strategies to correct errors that inevitably arise in physical quantum systems 1 . Quantum error-correction codes 2 present a way to reach this goal by encoding logical information redundantly into many physical qubits. A key challenge in implementing such codes is accurately decoding noisy syndrome information extracted from redundancy checks to obtain the correct encoded logical information. Here we develop a recurrent, transformer-based neural network that learns to decode the surface code, the leading quantum error-correction code 3 . Our decoder outperforms other state-of-the-art decoders on real-world data from Google’s Sycamore quantum processor for distance-3 and distance-5 surface codes 4 . On distances up to 11, the decoder maintains its advantage on simulated data with realistic noise including cross-talk and leakage, utilizing soft readouts and leakage information. After training on approximate synthetic data, the decoder adapts to the more complex, but unknown, underlying error distribution by training on a limited budget of experimental samples. Our work illustrates the ability of machine learning to go beyond human-designed algorithms by learning from data directly, highlighting machine learning as a strong contender for decoding in quantum computers. A recurrent, transformer-based neural network, called AlphaQubit, learns high-accuracy error decoding to suppress the errors that occur in quantum systems, opening the prospect of using neural-network decoders for real quantum hardware.

We have developed a reinforcement learning agent that often finds a minimal sequence of unknotting crossing changes for a knot diagram with up to 200 crossings, hence giving an upper bound on the unknotting number. We have used this to determine the unknotting number of 57k knots. We took diagrams of connected sums of such knots with oppositely signed signatures, where the summands were overlaid. The agent has found examples where several of the crossing changes in an unknotting collection of crossings result in hyperbolic knots. Based on this, we have shown that, given knots \\(K\\) and \\(K'\\) that satisfy some mild assumptions, there is a diagram of their connected sum and \\(u(K) + u(K')\\) unknotting crossings such that changing any one of them results in a prime knot. As a by-product, we have obtained a dataset of 2.6 million distinct hard unknot diagrams; most of them under 35 crossings. Assuming the additivity of the unknotting number, we have determined the unknotting number of 43 at most 12-crossing knots for which the unknotting number is unknown.

Quantum error-correction is a prerequisite for reliable quantum computation. Towards this goal, we present a recurrent, transformer-based neural network which learns to decode the surface code, the leading quantum error-correction code. Our decoder outperforms state-of-the-art algorithmic decoders on real-world data from Google's Sycamore quantum processor for distance 3 and 5 surface codes. On distances up to 11, the decoder maintains its advantage on simulated data with realistic noise including cross-talk, leakage, and analog readout signals, and sustains its accuracy far beyond the 25 cycles it was trained on. Our work illustrates the ability of machine learning to go beyond human-designed algorithms by learning from data directly, highlighting machine learning as a strong contender for decoding in quantum computers.

We have developed a reinforcement learning agent that often finds a minimal sequence of unknotting crossing changes for a knot diagram with up to 200 crossings, hence giving an upper bound on the unknotting number. We have used this to determine the unknotting number of 57k knots. We took diagrams of connected sums of such knots with oppositely signed signatures, where the summands were overlaid. The agent has found examples where several of the crossing changes in an unknotting collection of crossings result in hyperbolic knots. Based on this, we have shown that, given knots \\(K\\) and \\(K'\\) that satisfy some mild assumptions, there is a diagram of their connected sum and \\(u(K) + u(K')\\) unknotting crossings such that changing any one of them results in a prime knot. As a by-product, we have obtained a dataset of 2.6 million distinct hard unknot diagrams; most of them under 35 crossings. Assuming the additivity of the unknotting number, we have determined the unknotting number of 43 at most 12-crossing knots for which the unknotting number is unknown.

Fault-tolerant quantum computing will require error rates far below those achievable with physical qubits. Quantum error correction (QEC) bridges this gap, but depends on decoders being simultaneously fast, accurate, and scalable. This combination of requirements has not yet been met by a machine-learning decoder, nor by any decoder for promising resource-efficient codes such as the colour code. Here we introduce AlphaQubit 2, a neural-network decoder that achieves near-optimal logical error rates for both surface and colour codes at large scales under realistic noise. For the colour code, it is orders of magnitude faster than other high-accuracy decoders. For the surface code, we demonstrate real-time decoding faster than 1 microsecond per cycle up to distance 11 on current commercial accelerators with better accuracy than leading real-time decoders. These results support the practical application of a wider class of promising QEC codes, and establish a credible path towards high-accuracy, real-time neural decoding at the scales required for fault-tolerant quantum computation.

We apply a Transformer architecture, specifically BERT, to learn flexible and high quality molecular representations for drug discovery problems. We study the impact of using different combinations of self-supervised tasks for pre-training, and present our results for the established Virtual Screening and QSAR benchmarks. We show that: i) The selection of appropriate self-supervised task(s) for pre-training has a significant impact on performance in subsequent downstream tasks such as Virtual Screening. ii) Using auxiliary tasks with more domain relevance for Chemistry, such as learning to predict calculated molecular properties, increases the fidelity of our learnt representations. iii) Finally, we show that molecular representations learnt by our model `MolBert' improve upon the current state of the art on the benchmark datasets.

Janus kinases (JAKs) mediate responses to cytokines, hormones and growth factors in haematopoietic cells 1 , 2 . The JAK gene JAK2 is frequently mutated in the ageing haematopoietic system 3 , 4 and in haematopoietic cancers 5 . JAK2 mutations constitutively activate downstream signalling and are drivers of myeloproliferative neoplasm (MPN). In clinical use, JAK inhibitors have mixed effects on the overall disease burden of JAK2 -mutated clones 6 , 7 , prompting us to investigate the mechanism underlying disease persistence. Here, by in-depth phosphoproteome profiling, we identify proteins involved in mRNA processing as targets of mutant JAK2. We found that inactivation of YBX1 , a post-translationally modified target of JAK2, sensitizes cells that persist despite treatment with JAK inhibitors to apoptosis and results in RNA mis-splicing, enrichment for retained introns and disruption of the transcriptional control of extracellular signal-regulated kinase (ERK) signalling. In combination with pharmacological JAK inhibition, YBX1 inactivation induces apoptosis in JAK2-dependent mouse and primary human cells, causing regression of the malignant clones in vivo, and inducing molecular remission. This identifies and validates a cell-intrinsic mechanism whereby differential protein phosphorylation causes splicing-dependent alterations of JAK2–ERK signalling and the maintenance of JAK2 V617F malignant clones. Therapeutic targeting of YBX1-dependent ERK signalling in combination with JAK2 inhibition could thus eradicate cells harbouring mutations in JAK2 . Inhibition of YBX1, a downstream target of the Janus kinase JAK2, sensitizes myeloproliferative neoplasm cells to JAK and could provide a means to eradicate such cells in human haematopoietic cancers.

Objective Ribavirin improves treatment response to pegylated-interferon (PEG-IFN) in chronic hepatitis C but the mechanism remains controversial. We studied correlates of response and mechanism of action of ribavirin in treatment of hepatitis C. Design 70 treatment-naive patients were randomised to 4 weeks of ribavirin (1000–1200 mg/d) or none, followed by PEG-IFNα-2a and ribavirin at standard doses and durations. Patients were also randomised to a liver biopsy 24 h before or 6 h after starting PEG-IFN. Hepatic gene expression was assessed by microarray and interferon-stimulated gene (ISG) expression quantified by nCounter platform. Temporal changes in ISG expression were assessed by qPCR in peripheral-blood mononuclear cells (PBMC) and by serum levels of IP-10. Results After 4 weeks of ribavirin monotherapy, hepatitis C virus (HCV) levels decreased by 0.5±0.5 log10 (p=0.009 vs controls) and ALT by 33% (p<0.001). Ribavirin pretreatment, while modestly augmenting ISG induction by PEG-IFN, did not modify the virological response to subsequent PEG-IFN and ribavirin treatment. However, biochemical, but not virological, response to ribavirin monotherapy predicted response to subsequent combination treatment (rapid virological response, 71% in biochemical responders vs 22% non-responders, p=0.01; early virological response, 100% vs 68%, p=0.03; sustained virological response 83% vs 41%, p=0.053). Ribavirin monotherapy lowered serum IP-10 levels but had no effect on ISG expression in PBMC. Conclusions Ribavirin is a weak antiviral but its clinical effect seems to be mediated by a separate, indirect mechanism, which may act to reset IFN-responsiveness in HCV-infected liver.

The family of CRP proteins comprises three members, which are composed of two LIM domains separated by a long linker of more than 50 residues. We determined the structure of the muscle variant, MLP (CRP3), by nuclear magnetic resonance and show that the two LIM domains are independent of each other.