arXiv 论文速递

Multimodal large language models (MLLMs) are rapidly advancing clinical diagnosis, yet their adaptation pipelines remain anchored to accuracy-based objectives. Clinical data are he…

针对RLVR中LLM推理预测的信用分配问题，引入半事实信用增强策略优化，提升可验证奖励下推理训练的样本效率与稳定性。

Reinforcement learning with verifiable rewards (RLVR) has improved the reasoning capabilities of large language models (LLMs), yet their predictions remain sensitive to task-irrele…

揭示非侵入式脑机接口解码词句的"提升"大多可在无脑数据下复现，即模型依赖时间捷径而非神经信号，对脑机接口领域的方法论可信度构成根本性挑战。

We find that major reported improvements in decoding words from non-invasive brain recordings are largely reproducible without any brain data. In the influential work of d'Ascoli e…

Recent video generation is increasingly realistic and controllable, yet video editing remains less developed, particularly for precise local edits that must preserve the original s…

提出VideoMSN，用掩码孪生网络复用图像分类器高效学习视频时空表征，降低视频自监督预训练成本，兼具效率与性能。

We introduce VideoMSN, a Masked Siamese Network framework for efficient self-supervised spatio-temporal representation learning in videos. Instead of relying on heavy 3D architectu…

Differentially Private Stochastic Gradient Descent (DP-SGD) is a leading approach for privacy-preserving fine-tuning of large language models (LLMs). Many decoder-only LLMs employ …

Automating the search for effective harnesses is an important step toward enabling agents to recursively self-improve. Existing harness optimizations typically produce a single glo…

As interactive 3D worlds are increasingly used to study intelligent behavior, it becomes important to develop efficient pipelines for identifying anomalies in these simulated envir…

提出面向开放研究问题的多智能体证明发现框架，把前沿模型的推理能力组织成可协作的证明搜索流程，推动AI辅助数学发现。

We present Cogentic, a multi-agent harness for automated proof discovery on open research problems. While frontier language models can generate strong mathematical ideas in a singl…

Material generation should produce not only an appearance, but also the rules that construct it. We introduce MatLoom, a compact, layer-oriented language for text-to-material gener…

将循环Transformer与混合专家两条高效扩展路径统一，给出循环MoE的缩放律，为在固定参数下提升计算深度提供可预测的设计准则。

Looped transformers and Mixture-of-Experts (MoE) offer complementary routes to efficient scaling: recurrence increases computational depth at fixed parameters, while MoE sparsity e…

Lossy compression is widely used in Federated Learning (FL) but is generally treated as an error source, while conventional poisoning defenses inspect update geometry. In this work…