Jiaqi Lu

Tiantian Chen, Jiaqi Lu, Ying Shen et al.

Large Language Models (LLMs) have shown strong potential as conversational agents. Yet, their effectiveness remains limited by deficiencies in robust long-term memory, particularly in complex, long-term web-based services such as online emotional support. However, existing long-term dialogue benchmarks primarily focus on static and explicit fact retrieval, failing to evaluate agents in critical scenarios where user information is dispersed, implicit, and continuously evolving. To address this gap, we introduce ES-MemEval, a comprehensive benchmark that systematically evaluates five core memory capabilities: information extraction, temporal reasoning, conflict detection, abstention, and user modeling, in long-term emotional support settings, covering question answering, summarization, and dialogue generation tasks. To support the benchmark, we also propose EvoEmo, a multi-session dataset for personalized long-term emotional support that captures fragmented, implicit user disclosures and evolving user states. Extensive experiments on open-source long-context, commercial, and retrieval-augmented (RAG) LLMs show that explicit long-term memory is essential for reducing hallucinations and enabling effective personalization. At the same time, RAG improves factual consistency but struggles with temporal dynamics and evolving user states. These findings highlight both the potential and limitations of current paradigms and motivate more robust integration of memory and retrieval for long-term personalized dialogue systems.

Nic Fishman, Gokul Gowri, Tanush Kumar et al.

Many modern biological assays, including RNA sequencing, yield integer-valued counts that reflect the number of molecules detected. These measurements are often not at the desired resolution: while the unit of interest is typically a single cell, many measurement technologies produce counts aggregated over sets of cells. Although recent generative frameworks such as diffusion and flow matching have been extended to non-Euclidean and discrete settings, it remains unclear how best to model integer-valued data or how to systematically deconvolve aggregated observations. We introduce Count Bridges, a stochastic bridge process on the integers that provides an exact, tractable analogue of diffusion-style models for count data, with closed-form conditionals for efficient training and sampling. We extend this framework to enable direct training from aggregated measurements via an Expectation-Maximization-style approach that treats unit-level counts as latent variables. We demonstrate state-of-the-art performance on integer distribution matching benchmarks, comparing against flow matching and discrete flow matching baselines across various metrics. We then apply Count Bridges to two large-scale problems in biology: modeling single-cell gene expression data at the nucleotide resolution, with applications to deconvolving bulk RNA-seq, and resolving multicellular spatial transcriptomic spots into single-cell count profiles. Our methods offer a principled foundation for generative modeling and deconvolution of biological count data across scales and modalities.

Mianjun Xiao, Peng Song, Yulong Liu et al.

Finite element method (FEM) is widely used in high-temperature superconducting (HTS) magnets, but its computational cost increases with magnet size and becomes time-consuming for meter-scale magnets, especially when multi-physics couplings are considered, which limits the fast design of large-scale REBCO magnet systems. In this work, a surrogate model based on a fully connected residual neural network (FCRN) is developed to predict the space-time current density distribution in REBCO solenoids. Training datasets were generated from FEM simulations with varying numbers of turns and pancakes. The results demonstrate that, for deeper networks, the FCRN architecture achieves better convergence than conventional fully connected network (FCN), with the configuration of 12 residual blocks and 256 neurons per layer providing the most favorable balance between training accuracy and generalization capability. Extrapolation studies show that the model can reliably predict magnetization losses for up to 50% beyond the training range, with maximum errors below 10%. The surrogate model achieves predictions several orders of magnitude faster than FEM and still remains advantageous when training costs are included. These results indicate that the proposed FCRN-based surrogate model provides both accuracy and efficiency, offering a promising tool for the rapid analysis of large-scale HTS magnets.

Yilun Chen, Jiaqi Lu

We characterize a joint CLT of the number of pulls and the sample mean reward of the arms in a stochastic two-armed bandit environment under UCB algorithms. Several implications of this result are in place: (1) a nonstandard CLT of the number of pulls hence pseudo-regret that smoothly interpolates between a standard form in the large arm gap regime and a slow-concentration form in the small arm gap regime, and (2) a heuristic derivation of the sample bias up to its leading order from the correlation between the number of pulls and sample means. Our analysis framework is based on a novel perturbation analysis, which is of broader interest on its own.