The brain's diverse spatiotemporal activity patterns are fundamental to cognition and consciousness, yet how these macroscopic dynamics emerge from microscopic neural circuitry remains a critical challenge. We take a step in this direction by developing a spatially extended neural network model integrated with a spectral theory of its connectivity matrix. Our theory quantitatively demonstrates how local structural parameters, such as E/I neuron projection ranges, connection strengths, and density determine distinct features of the eigenvalue spectrum, specifically outlier eigenvalues and a bulk disk. These spectral signatures, in turn, precisely predict the network's emergent global dynamical regime, encompassing asynchronous states, synchronous states, oscillations, localized activity bumps, traveling waves, and chaos. Motivated by observations of shifting cortical dynamics in mice across arousal states, our framework not only provides a possible explanation for repertoire of behaviors but also offers a principled starting point for inferring underlying effective connectivity changes from macroscopic brain activity. By mechanistically linking neural structure to dynamics, this work advances a principled framework for dissecting how large-scale activity patterns—central to cognition and open questions in consciousness research—arise from, and constrain, local circuitry. The implementation code is available at https://github.com/huang-yh20/spatial-linear-project.

Electroencephalogram (EEG)-based emotion decoding can objectively quantify people's emotional state and has broad application prospects in human-computer interaction and early detection of emotional disorders. Recently emerging deep learning architectures have significantly improved the performance of EEG emotion decoding. However, existing methods still fall short of fully capturing the complex spatiotemporal dynamics of neural signals, which are crucial for representing emotion processing. This study proposes a Dynamic-Attention-based EEG State Transition (DAEST) modeling method to characterize EEG spatiotemporal dynamics. The model extracts spatiotemporal components of EEG that represent multiple parallel neural processes and estimates dynamic attention weights on these components to capture transitions in brain states. The model is optimized within a contrastive learning framework for cross-subject emotion recognition. The proposed method achieved state-of-the-art performance on three publicly available datasets: FACED, SEED, and SEED-V. It achieved$81.7\pm 4.3\%$accuracy in the binary classification of positive and negative emotions and$67.9\pm 7.3\%$in nine-class discrete emotion classification on the FACED dataset,$88.1\pm 3.6\%$in the three-class classification of positive, negative, and neutral emotions on the SEED dataset, and$73.6\pm 12.7\%$in five-class discrete emotion classification on the SEED-V dataset. The learned EEG spatiotemporal patterns and dynamic transition properties offer valuable insights into neural dynamics underlying emotion processing.

Nowadays, open-source large language models like LLaMA have emerged. Recent developments have incorporated supervised fine-tuning (SFT) and reinforcement learning fine-tuning (RLFT) to align these models with human goals. However, SFT methods treat all training data with mixed quality equally, while RLFT methods require high-quality pairwise or ranking-based preference data. In this study, we present a novel framework, named OpenChat, to advance open-source language models with mixed-quality data. Specifically, we consider the general SFT training data, consisting of a small amount of expert data mixed with a large proportion of sub-optimal data, without any preference labels. We propose the C(onditioned)-RLFT, which regards different data sources as coarse-grained reward labels and learns a class-conditioned policy to leverage complementary data quality information. Interestingly, the optimal policy in C-RLFT can be easily solved through single-stage, RL-free supervised learning, which is lightweight and avoids costly human preference labeling. Through extensive experiments on three standard benchmarks, our openchat-13b fine-tuned with C-RLFT achieves the highest average performance among all 13b open-source language models. Moreover, we use AGIEval to validate the model generalization performance, in which only openchat-13b surpasses the base model. Finally, we conduct a series of analyses to shed light on the effectiveness and robustness of OpenChat. Our code, data, and models are publicly available at https://github.com/imoneoi/openchat and https://huggingface.co/openchat.

Named entity recognition is a fundamental task in natural language processing. Based on the sequence labeling paradigm for flat named entity recognition, multiple methods have been developed to handle the nested structures. However, they either require fixed recognition order or introduce complex hypergraphs. To tackle this problem, we propose a novel model named Local Hypergraph Builder Network (LHBN) that builds multiple simpler local hypergraphs to capture named entities instead of a single complex full-size hypergraph. The proposed model has three main properties: (1) The named entities that share boundaries are captured in the same local hypergraph. (2) The boundary information is enhanced by building local hypergraphs. (3) The hypergraphs can be built bidirectionally to take advantage of the identification direction preference of different named entities. Experiments illustrate that our model outperforms previous state-of-the-art methods on four widely used nested named entity recognition datasets: ACE04, ACE05, GENIA, and KBP17. The code is available at https://github.com/yanyk13/local-hypergraph-building-network.git.

Recurrent spiking neural networks (RSNNs) hold great potential for advancing artificial general intelligence, as they draw inspiration from the biological nervous system and show promise in modeling complex dynamics. However, the widely-used surrogate gradient-based training methods for RSNNs are inherently inaccurate and unfriendly to neuromorphic hardware. To address these limitations, we propose the evolving connectivity (EC) framework, an inference-only method for training RSNNs. The EC framework reformulates weight-tuning as a search into parameterized connection probability distributions, and employs Natural Evolution Strategies (NES) for optimizing these distributions. Our EC framework circumvents the need for gradients and features hardware-friendly characteristics, including sparse boolean connections and high scalability. We evaluate EC on a series of standard robotic locomotion tasks, where it achieves comparable performance with deep neural networks and outperforms gradient-trained RSNNs, even solving the complex 17-DoF humanoid task. Additionally, the EC framework demonstrates a two to three fold speedup in efficiency compared to directly evolving parameters. By providing a performant and hardware-friendly alternative, the EC framework lays the groundwork for further energy-efficient applications of RSNNs and advances the development of neuromorphic devices. Our code is publicly available at https://github.com/imoneoi/EvolvingConnectivity.

EEG signals have been reported to be informative and reliable for emotion recognition in recent years. However, the inter-subject variability of emotion-related EEG signals still poses a great challenge for the practical applications of EEG-based emotion recognition. Inspired by recent neuroscience studies on inter-subject correlation, we proposed a Contrastive Learning method for Inter-Subject Alignment (CLISA) to tackle the cross-subject emotion recognition problem. Contrastive learning was employed to minimize the inter-subject differences by maximizing the similarity in EEG signkal representations across subjects when they received the same emotional stimuli in contrast to different ones. Specifically, a convolutional neural network was applied to learn inter-subject aligned spatiotemporal representations from EEG time series in contrastive learning. The aligned representations were subsequently used to extract differential entropy features for emotion classification. CLISA achieved state-of-the-art cross-subject emotion recognition performance on our THU-EP dataset with 80 subjects and the publicly available SEED dataset with 15 subjects. It could generalize to unseen subjects or unseen emotional stimuli in testing. Furthermore, the spatiotemporal representations learned by CLISA could provide insights into the neural mechanisms of human emotion processing.

Abstract reasoning, as one of the hallmarks of human intelligence, involves collecting information, identifying abstract rules, and applying the rules to solve new problems. Although neural networks have achieved human-level performances in several tasks, the abstract reasoning techniques still far lag behind due to the complexity of learning and applying the logic rules, especially in an unsupervised manner. In this work, we propose a novel framework, ARII, that learns rule representations for Abstract Reasoning via Internal Inferences. The key idea is to repeatedly apply a rule to different instances in hope of having a comprehensive understanding (i.e., representations) of the rule. Specifically, ARII consists of a rule encoder, a reasoner, and an internal referrer. Based on the representations produced by the rule encoder, the reasoner draws the conclusion while the referrer performs internal inferences to regularize rule representations to be robust and generalizable. We evaluate ARII on two benchmark datasets, including PGM and I-RAVEN. We observe that ARII achieves new state-of-the-art records on the majority of the reasoning tasks, including most of the generalization tests in PGM. Our codes are available at https://github.com/Zhangwenbo0324/ARII.

Self-supervised learning has gradually emerged as a powerful technique for graph representation learning. However, transferable, generalizable, and robust representation learning on graph data still remains a challenge for pre-training graph neural networks. In this paper, we propose a simple and effective self-supervised pre-training strategy, named Pairwise Half-graph Discrimination (PHD), that explicitly pre-trains a graph neural network at graph-level. PHD is designed as a simple binary classification task to discriminate whether two half-graphs come from the same source. Experiments demonstrate that the PHD is an effective pre-training strategy that offers comparable or superior performance on 13 graph classification tasks compared with state-of-the-art strategies, and achieves notable improvements when combined with node-level strategies. Moreover, the visualization of learned representation revealed that PHD strategy indeed empowers the model to learn graph-level knowledge like the molecular scaffold. These results have established PHD as a powerful and effective self-supervised learning strategy in graph-level representation learning.

Translation elongation is a crucial phase during protein biosynthesis. In this study, we develop a novel deep reinforcement learning-based framework, named Riboexp, to model the determinants of the uneven distribution of ribosomes on mRNA transcripts during translation elongation. In particular, our model employs a policy network to perform a context-dependent feature selection in the setting of ribosome density prediction. Our extensive tests demonstrated that Riboexp can significantly outperform the state-of-the-art methods in predicting ribosome density by up to 5.9% in terms of per-gene Pearson correlation coefficient on the datasets from three species. In addition, Riboexp can indicate more informative sequence features for the prediction task than other commonly used attribution methods in deep learning. In-depth analyses also revealed the meaningful biological insights generated by the Riboexp framework. Moreover, the application of Riboexp in codon optimization resulted in an increase of protein production by around 31% over the previous state-of-the-art method that models ribosome density. These results have established Riboexp as a powerful and useful computational tool in the studies of translation dynamics and protein synthesis. Availability: The data and code of this study are available on GitHub: https://github.com/Liuxg16/Riboexp. Contact:[email protected]; [email protected].

MOTIVATION: Computational methods accelerate drug discovery and play an important role in biomedicine, such as molecular property prediction and compound-protein interaction (CPI) identification. A key challenge is to learn useful molecular representation. In the early years, molecular properties are mainly calculated by quantum mechanics or predicted by traditional machine learning methods, which requires expert knowledge and is often labor-intensive. Nowadays, graph neural networks have received significant attention because of the powerful ability to learn representation from graph data. Nevertheless, current graph-based methods have some limitations that need to be addressed, such as large-scale parameters and insufficient bond information extraction. RESULTS: In this study, we proposed a graph-based approach and employed a novel triplet message mechanism to learn molecular representation efficiently, named triplet message networks (TrimNet). We show that TrimNet can accurately complete multiple molecular representation learning tasks with significant parameter reduction, including the quantum properties, bioactivity, physiology and CPI prediction. In the experiments, TrimNet outperforms the previous state-of-the-art method by a significant margin on various datasets. Besides the few parameters and high prediction accuracy, TrimNet could focus on the atoms essential to the target properties, providing a clear interpretation of the prediction tasks. These advantages have established TrimNet as a powerful and useful computational tool in solving the challenging problem of molecular representation learning. AVAILABILITY: The quantum and drug datasets are available on the website of MoleculeNet: http://moleculenet.ai. The source code is available in GitHub: https://github.com/yvquanli/trimnet. CONTACT: [email protected], [email protected].

How to produce expressive molecular representations is a fundamental challenge in artificial intelligence-driven drug discovery. Graph neural network (GNN) has emerged as a powerful technique for modeling molecular data. However, previous supervised approaches usually suffer from the scarcity of labeled data and poor generalization capability. Here, we propose a novel molecular pre-training graph-based deep learning framework, named MPG, that learns molecular representations from large-scale unlabeled molecules. In MPG, we proposed a powerful GNN for modelling molecular graph named MolGNet, and designed an effective self-supervised strategy for pre-training the model at both the node and graph-level. After pre-training on 11 million unlabeled molecules, we revealed that MolGNet can capture valuable chemical insights to produce interpretable representation. The pre-trained MolGNet can be fine-tuned with just one additional output layer to create state-of-the-art models for a wide range of drug discovery tasks, including molecular properties prediction, drug-drug interaction and drug-target interaction, on 14 benchmark datasets. The pre-trained MolGNet in MPG has the potential to become an advanced molecular encoder in the drug discovery pipeline.

Modeling the impact of amino acid mutations on protein-protein interaction plays a crucial role in protein engineering and drug design. In this study, we develop GeoPPI, a novel structure-based deep-learning framework to predict the change of binding affinity upon mutations. Based on the three-dimensional structure of a protein, GeoPPI first learns a geometric representation that encodes topology features of the protein structure via a self-supervised learning scheme. These representations are then used as features for training gradient-boosting trees to predict the changes of protein-protein binding affinity upon mutations. We find that GeoPPI is able to learn meaningful features that characterize interactions between atoms in protein structures. In addition, through extensive experiments, we show that GeoPPI achieves new state-of-the-art performance in predicting the binding affinity changes upon both single- and multi-point mutations on six benchmark datasets. Moreover, we show that GeoPPI can accurately estimate the difference of binding affinities between a few recently identified SARS-CoV-2 antibodies and the receptor-binding domain (RBD) of the S protein. These results demonstrate the potential of GeoPPI as a powerful and useful computational tool in protein design and engineering. Our code and datasets are available at: https://github.com/Liuxg16/GeoPPI.

We propose UPSA, a novel approach that accomplishes Unsupervised Paraphrasing by Simulated Annealing.We model paraphrase generation as an optimization problem and propose a sophisticated objective function, involving semantic similarity, expression diversity, and language fluency of paraphrases.UPSA searches the sentence space towards this objective by performing a sequence of local edits.We evaluate our approach on various datasets, namely, Quora, Wikianswers, MSCOCO, and Twitter.Extensive results show that UPSA achieves the state-of-the-art performance compared with previous unsupervised methods in terms of both automatic and human evaluations.Further, our approach outperforms most existing domain-adapted supervised models, showing the generalizability of UPSA. 1

Deep generative modeling has achieved many successes for continuous data generation, such as producing realistic images and controlling their properties (e.g., styles). However, the development of generative modeling techniques for optimizing discrete data, such as sequences or strings, still lags behind largely due to the challenges in modeling complex and long-range constraints, including both syntax and semantics, in discrete structures. In this paper, we formulate the sequence optimization task as a chance-constrained optimization problem. The key idea is to enforce a high probability of generating valid sequences and also optimize the property of interest. We propose a novel minimax algorithm to simultaneously tighten a bound of the valid chance and optimize the expected property. Extensive experimental results in three domains demonstrate the superiority of our approach over the existing sequence optimization methods.

With the increasing demand of applications that have two different output voltages, the single-input dual-output (SIDO) converter with fewer components is becoming the cost-effective option instead of employing two single-input single-output converters. However, the cross-regulation of different outputs is still a challenge in SIDO converters design. To save cost and obtain improved cross-regulation performance, a novel SIDO converter consisting of three active switches is proposed in this article. Owing to the use of a voltage multiplier circuit, a high step-up voltage conversion ratio is achieved with relatively low voltage stress of switches. Thanks to the independent power flow and two control variables, the cross-regulation performance is improved, and simultaneous buck as well as boost output voltages is realized. Additionally, all switches can achieve zero-voltage switching operation, which contributes to a significant switching loss reduction. The operation characteristics, design considerations, and control strategy of the proposed converter are analyzed. To verify the theoretical analysis and measure the power efficiency, a prototype circuit with 48 V input and 24 V/2 A, 200 V/1 A outputs is built.

Segmenting human left ventricle (LV) in magnetic resonance imaging images and calculating its volume are important for diagnosing cardiac diseases. The latter task became the topic of the Second Annual Data Science Bowl organized by Kaggle. The dataset consisted of a large number of cases with only systole and diastole volume labels. We designed a system based on neural networks to solve this problem. It began with a detector to detect the regions of interest (ROI) containing LV chambers. Then a deep neural network named hypercolumns fully convolutional network was used to segment LV in ROI. The 2-D segmentation results were integrated across different images to estimate the volume. With ground-truth volume labels, this model was trained end-to-end. To improve the result, an additional dataset with only segmentation labels was used. The model was trained alternately on these two tasks. We also proposed a variance estimation method for the final prediction. Our algorithm ranked the fourth on the test set in this competition.

Automatic diagnosing lung cancer from computed tomography scans involves two steps: detect all suspicious lesions (pulmonary nodules) and evaluate the whole-lung/pulmonary malignancy. Currently, there are many studies about the first step, but few about the second step. Since the existence of nodule does not definitely indicate cancer, and the morphology of nodule has a complicated relationship with cancer, the diagnosis of lung cancer demands careful investigations on every suspicious nodule and integration of information of all nodules. We propose a 3-D deep neural network to solve this problem. The model consists of two modules. The first one is a 3-D region proposal network for nodule detection, which outputs all suspicious nodules for a subject. The second one selects the top five nodules based on the detection confidence, evaluates their cancer probabilities, and combines them with a leaky noisy-OR gate to obtain the probability of lung cancer for the subject. The two modules share the same backbone network, a modified U-net. The overfitting caused by the shortage of the training data is alleviated by training the two modules alternately. The proposed model won the first place in the Data Science Bowl 2017 competition.

In early years, text classification is typically accomplished by feature-based classifiers; recently, neural networks, as powerful classifiers, make it possible to work with raw input as the text stands. In this paper, we propose a novel framework, Jumper, inspired by the cognitive process of text reading, that models text classification as a sequential decision process. Basically, Jumper is a neural system that can scan a piece of text sequentially and make classification decision at the time it chooses. Both the classification and when to make the classification are part of the decision process which are controlled by the policy net and trained with reinforcement learning to maximize the overall classification accuracy. Experimental results show that a properly trained Jumper has the following properties: (1) It can make decisions whenever the evidence is enough, therefore reducing the total text reading by 30~40% and often finding the key rationale of prediction. (2) It can achieve classification accuracy better or comparable to state-of-the-art model in several benchmark and industrial datasets.

It has been debated whether kinematic features, such as the number of peaks or decomposed submovements in a velocity profile, indicate the number of discrete motor impulses or result from a continuous control process. The debate is particularly relevant for tasks involving target perturbation, which can alter movement kinematics. To simulate such tasks, finite-horizon models require two preset movement durations to compute two control policies before and after the perturbation. Another model employs infinite- and finite-horizon formulations to determine, respectively, movement durations and control policies, which are updated every time step. We adopted an infinite-horizon optimal feedback control model that, unlike previous approaches, does not preset movement durations or use multiple control policies. It contains both control-dependent and independent noises in system dynamics, state-dependent and independent noises in sensory feedbacks, and different delays and noise levels for visual and proprioceptive feedbacks. We analytically derived an optimal solution that can be applied continuously to move an effector toward a target regardless of whether, when, or where the target jumps. This single policy produces different numbers of peaks and "submovements" in velocity profiles for different conditions and trials. Movements that are slower or perturbed later appear to have more submovements. The model is also consistent with the observation that subjects can perform the perturbation task even without detecting the target jump or seeing their hands during reaching. Finally, because the model incorporates Weber's law via a state representation relative to the target, it explains why initial and terminal visual feedback are, respectively, less and more effective in improving end-point accuracy. Our work suggests that the number of peaks or submovements in a velocity profile does not necessarily reflect the number of motor impulses and that the difference between initial and terminal feedback does not necessarily imply a transition between open- and closed-loop strategies.

In recent years, convolutional neural network (CNN) based methods have achieved great success in a large number of applications and have been among the most powerful and widely used techniques in computer vision. However, CNN-based methods are com-putational-intensive and resource-consuming, and thus are hard to be integrated into embedded systems such as smart phones, smart glasses, and robots. FPGA is one of the most promising platforms for accelerating CNN, but the limited bandwidth and on-chip memory size limit the performance of FPGA accelerator for CNN.