Biological sequences, including RNAs and proteins, share similarities with natural languages, enabling the application of advanced language models to various biological tasks. However, due to its flexibility and lack of experimental data, RNA is a particularly challenging biological ``language'' compared to other biological sequences like proteins. RNA multiple sequence alignments (MSAs), which align evolutionarily related RNA sequences, can greatly enhance RNA biology modeling, as evidenced by their significant roles in structure prediction and function annotation. This raises the question of whether RNA MSAs can also benefit RNA design, which remains unexplored. This paper introduces RMSAGen, a model comprising RMSA-Encoder and RMSA-Decoder, that leverages MSAs to design functional RNA sequences. RMSA-Encoder effectively extracts MSA features, enhancing performance in functional prediction and solvent accessibility prediction tasks and supporting RMSA-Decoder in accurate RNA generation. RMSAGen can design RNA sequences that effectively bind to target RNA-binding proteins, and the design performance improves with an increasing number of sequences. In addition, the ribozymes designed with structural features by RMSAGen show strong computational metrics and exhibit biological activity during gel electrophoresis. These results highlight the effectiveness of RMSAGen, establishing it as a powerful tool and a new direction for RNA design.

Inverse Protein Folding (IPF) is a critical subtask in the field of protein design, aiming to engineer amino acid sequences capable of folding correctly into a specified three-dimensional (3D) conformation. Although substantial progress has been achieved in recent years, existing methods generally rely on either backbone coordinates or molecular surface features alone, which restricts their ability to fully capture the complex chemical and geometric constraints necessary for precise sequence prediction. To address this limitation, we present DS-ProGen, a dual-structure deep language model for functional protein design, which integrates both backbone geometry and surface-level representations. By incorporating backbone coordinates as well as surface chemical and geometric descriptors into a next-amino-acid prediction paradigm, DS-ProGen is able to generate functionally relevant and structurally stable sequences while satisfying both global and local conformational constraints. On the PRIDE dataset, DS-ProGen attains the current state-of-the-art recovery rate of 61.47%, demonstrating the synergistic advantage of multi-modal structural encoding in protein design. Furthermore, DS-ProGen excels in predicting interactions with a variety of biological partners, including ligands, ions, and RNA, confirming its robust functional retention capabilities.

Large Language Models (LLMs) have demonstrated remarkable in-context learning (ICL) capabilities. In this study, we explore a surprising phenomenon related to ICL: LLMs can perform multiple, computationally distinct ICL tasks simultaneously, during a single inference call, a capability we term task superposition". We provide empirical evidence of this phenomenon across various LLM families and scales and show that this phenomenon emerges even if we train the model to in-context learn one task at a time. We offer theoretical explanations that this capability is well within the expressive power of transformers. We also explore how LLMs internally compose task vectors during superposition. Furthermore, we show that larger models can solve more ICL tasks in parallel, and better calibrate their output distribution. Our findings offer insights into the latent capabilities of LLMs, further substantiate the perspective of "LLMs as superposition of simulators", and raise questions about the mechanisms enabling simultaneous task execution.

In-context learning (ICL) exhibits dual operating modes: ***task learning***, i.e., acquiring a new skill from in-context samples, and ***task retrieval***, i.e., locating and activating a relevant pretrained skill. Recent theoretical work proposes various mathematical models to analyze ICL, but they cannot fully explain the duality. In this work, we analyze a generalized probabilistic model for pretraining data, obtaining a quantitative understanding of the two operating modes of ICL. Leveraging our analysis, we provide the first explanation of an unexplained phenomenon observed with real-world large language models (LLMs). Under some settings, the ICL risk initially increases and then decreases with more in-context examples. Our analysis offers a plausible explanation for this "early ascent" phenomenon: a limited number of in-context samples may lead to the retrieval of an incorrect skill, thereby increasing the risk, which will eventually diminish as task learning takes effect with more in-context samples. We also analyze ICL with biased labels, e.g., zero-shot ICL, where in-context examples are assigned random labels, and predict the bounded efficacy of such approaches. We corroborate our analysis and predictions with extensive experiments with Transformers and LLMs.

Recent studies on pre-trained vision/language models have demonstrated the practical benefit of a new, promising solution-building paradigm in AI where models can be pre-trained on broad data describing a generic task space and then adapted successfully to solve a wide range of downstream tasks, even when training data is severely limited (e.g., in zero- or few-shot learning scenarios). Inspired by such progress, we investigate in this paper the possibilities and challenges of adapting such a paradigm to the context of recommender systems, which is less investigated from the perspective of pre-trained model. In particular, we propose to develop a generic recommender that captures universal interaction patterns by training on generic user-item interaction data extracted from different domains, which can then be fast adapted to improve few-shot learning performance in unseen new domains (with limited data).

MOTIVATION: RNA design shows growing applications in synthetic biology and therapeutics, driven by the crucial role of RNA in various biological processes. A fundamental challenge is to find functional RNA sequences that satisfy given structural constraints, known as the inverse folding problem. Computational approaches have emerged to address this problem based on secondary structures. However, designing RNA sequences directly from 3D structures is still challenging, due to the scarcity of data, the nonunique structure-sequence mapping, and the flexibility of RNA conformation. RESULTS: In this study, we propose RiboDiffusion, a generative diffusion model for RNA inverse folding that can learn the conditional distribution of RNA sequences given 3D backbone structures. Our model consists of a graph neural network-based structure module and a Transformer-based sequence module, which iteratively transforms random sequences into desired sequences. By tuning the sampling weight, our model allows for a trade-off between sequence recovery and diversity to explore more candidates. We split test sets based on RNA clustering with different cut-offs for sequence or structure similarity. Our model outperforms baselines in sequence recovery, with an average relative improvement of 11% for sequence similarity splits and 16% for structure similarity splits. Moreover, RiboDiffusion performs consistently well across various RNA length categories and RNA types. We also apply in silico folding to validate whether the generated sequences can fold into the given 3D RNA backbones. Our method could be a powerful tool for RNA design that explores the vast sequence space and finds novel solutions to 3D structural constraints. AVAILABILITY AND IMPLEMENTATION: The source code is available at https://github.com/ml4bio/RiboDiffusion.

To ensure the safety of Autonomous Vehicles (AVs), thorough testing across virtual simulation environments, closed facilities, and public roads is essential. Scenario-based testing stands out as a crucial method for evaluating AVs, with a key focus on constructing appropriate testing scenarios. Given the vulnerability of Powered Two-Wheelers (PTWs) riders, it is essential to investigate typical and representative car-to-PTWs crash scenarios and validate AV system safety performance in such situations. This study introduces a new method for generating high-risk scenarios by extracting typical testing scenarios from real-world crashes, thereby enhancing the realism of testing conditions. To evaluate the safety performance of AV systems, a crash-based testing approach is proposed. First, 222 car-to-PTWs crashes were extracted from the China In-depth Mobility Safety Study-Traffic Accident (CIMSS-TA) database and the pre-crash trajectories of crash-involved parties were accurately obtained by reconstructing the crashes case-by-case. Second, utilizing the$k$-medoids algorithm based on the Hausdorff distance to cluster these trajectories, six typical pre-crash trajectory clusters were extracted. Third, six sets of high-risk scenarios were generated using parameter discretization and combination testing based on the cluster centers. Finally, we conducted safety testing on the black-box automated driving system, Baidu Apollo, using the SVL Simulator virtual simulation platform. We evaluated its performance by subjecting it to the six sets of high-risk scenarios generated in this study. The experimental results demonstrate that Apollo can operate safely in most high-risk scenarios, indicating that automated driving systems can handle crashes that some human drivers cannot avoid, thereby improve traffic safety.

Fine-tuning pretrained language models (LMs) without making any architectural changes has become a norm for learning various language downstream tasks. However, for non-language downstream tasks, a common practice is to employ task-specific designs for input, output layers, and loss functions. For instance, it is possible to fine-tune an LM into an MNIST classifier by replacing the word embedding layer with an image patch embedding layer, the word token output layer with a 10-way output layer, and the word prediction loss with a 10-way classification loss, respectively. A natural question arises: Can LM fine-tuning solve non-language downstream tasks without changing the model architecture or loss function? To answer this, we propose Language-Interfaced Fine-Tuning (LIFT) and study its efficacy and limitations by conducting an extensive empirical study on a suite of non-language classification and regression tasks. LIFT does not make any changes to the model architecture or loss function, and it solely relies on the natural language interface, enabling "no-code machine learning with LMs." We find that LIFT performs comparably well across a wide range of low-dimensional classification and regression tasks, matching the performances of the best baselines in many cases, especially for the classification tasks. We also report experimental results on the fundamental properties of LIFT, including inductive bias, robustness, and sample complexity. We also analyze the effect of pretraining on LIFT and a few properties/techniques specific to LIFT, e.g., context-aware learning via appropriate prompting, calibrated predictions, data generation, and two-stage fine-tuning. Our code is available at https://github.com/UW-Madison-Lee-Lab/LanguageInterfacedFineTuning.

Crowd flow prediction is of great importance in a wide range of applications from urban planning, traffic control to public safety. It aims at predicting the inflow (the traffic of crowds entering a region in a given time interval) and outflow (the traffic of crowds leaving a region for other places) of each region in the city with knowing the historical flow data. In this article, we propose DeepSTN+, a deep learning-based convolutional model, to predict crowd flows in the metropolis. First, DeepSTN+ employs the ConvPlus structure to model the long-range spatial dependence among crowd flows in different regions. Further, PoI distributions and time factor are combined to express the effect of location attributes to introduce prior knowledge of the crowd movements. Finally, we propose a temporal attention-based fusion mechanism to stabilize the training process, which further improves the performance. Extensive experimental results based on four real-life datasets demonstrate the superiority of our model, i.e., DeepSTN+ reduces the error of the crowd flow prediction by approximately 10%–21% compared with the state-of-the-art baselines.

Out-of-distribution (OOD) detection is essential to prevent anomalous inputs from causing a model to fail during deployment. While improved OOD detection methods have emerged, they often rely on the final layer outputs and require a full feedforward pass for any given input. In this paper, we propose a novel framework, multi-level out-of-distribution detection (MOOD), which exploits intermediate classifier outputs for dynamic and efficient OOD inference. We explore and establish a direct relationship between the OOD data complexity and optimal exit level, and show that easy OOD examples can be effectively detected early without propagating to deeper layers. At each exit, the OOD examples can be distinguished through our proposed adjusted energy score, which is both empirically and theoretically suitable for networks with multiple classifiers. We extensively evaluate MOOD across 10 OOD datasets spanning a wide range of complexities. Experiments demonstrate that MOOD achieves up to 71.05% computational reduction in inference, while maintaining competitive OOD detection performance.

Population flow prediction is one of the most fundamental components in many applications from urban management to transportation schedule. It is challenging due to the complicated spatial-temporal correlation.While many studies have been done in recent years, they fail to simultaneously and effectively model the spatial correlation and temporal variations among population flows. In this paper, we propose Convolution based Sequential and Cross Network (CSCNet) to solve them. On the one hand, we design a CNN based sequential structure with progressively merging the flow features from different time in different CNN layers to model the spatial-temporal information simultaneously. On the other hand, we make use of the transition flow as the proxy to efficiently and explicitly capture the dynamic correlation between different types of population flows. Extensive experiments on 4 datasets demonstrate that CSCNet outperforms the state-of-the-art baselines by reducing the prediction error around 7.7%∼10.4%.

Crowd flow prediction is of great importance in a wide range of applications from urban planning, traffic control to public safety. It aims to predict the inflow (the traffic of crowds entering a region in a given time interval) and outflow (the traffic of crowds leaving a region for other places) of each region in the city with knowing the historical flow data. In this paper, we propose DeepSTN+, a deep learning-based convolutional model, to predict crowd flows in the metropolis. First, DeepSTN+ employs the ConvPlus structure to model the longrange spatial dependence among crowd flows in different regions. Further, PoI distributions and time factor are combined to express the effect of location attributes to introduce prior knowledge of the crowd movements. Finally, we propose an effective fusion mechanism to stabilize the training process, which further improves the performance. Extensive experimental results based on two real-life datasets demonstrate the superiority of our model, i.e., DeepSTN+ reduces the error of the crowd flow prediction by approximately 8%∼13% compared with the state-of-the-art baselines.