5622 papers
The paper introduces MrBERT, a family of efficient, open-source multilingual encoders built on the ModernBERT architecture that achieves state-of-the-art performance in specific languages and specialized domains while leveraging Matryoshka Representation Learning to reduce inference and storage costs.
The paper introduces AVDE, a lightweight and efficient autoregressive framework that leverages contrastive learning and multi-scale token prediction to decode EEG signals into coherent images, outperforming state-of-the-art methods with significantly fewer parameters while mimicking the hierarchical nature of human visual perception.
CeRA overcomes the linear performance ceiling of Low-Rank Adaptation (LoRA) in complex reasoning tasks by introducing a weight-level parallel adapter with SiLU gating and structural dropout to induce manifold expansion, thereby achieving superior spectral efficiency and preventing rank collapse.
This paper addresses the scarcity of expert textual relevance labels in large-scale app store search by leveraging a specialized, fine-tuned LLM to generate millions of high-quality labels, which, when used to augment the production ranker, significantly improves both offline metrics and real-world conversion rates, particularly for tail queries lacking reliable behavioral data.
This paper introduces the first end-to-end differentiable optical particle detector simulator that unifies simulation, calibration, and reconstruction into a single gradient-based framework, demonstrating improved accuracy, speed, and flexibility for analyzing large-scale neutrino detectors compared to traditional methods.
This paper introduces Attn-QAT, the first systematic 4-bit quantization-aware training framework for attention mechanisms that ensures stable FP4 training and inference by matching low-precision recomputation in the backward pass and correcting implicit precision assumptions, thereby eliminating quality drops and delivering up to 1.5x speedup on FP4-capable GPUs without relying on outlier-mitigation heuristics.
This paper introduces a new class of non-equal volume partitions that achieve a lower expected star discrepancy and improved upper bounds compared to classical jittered sampling, thereby providing a theoretical foundation for their use in high-dimensional numerical integration.
This paper introduces a reproducible, scalable framework for evaluating multimodal models on ECG signals by decomposing reasoning into "Perception" (verified via code generation) and "Deduction" (verified via retrieval against clinical criteria) to address the limitations of existing manual or superficial evaluation methods.
This paper proposes a tractable two-layer framework combining a Hidden Markov Model for inferring rival energy states and a Deep Q-Network for decision-making to optimize 2026 Formula 1 energy strategies under partial observability, specifically addressing the "counter-harvest trap" where opponents deliberately mask their deployment signals.
The TCG CREST system achieved a sixth-place ranking in the DISPLACE-M challenge's speaker diarization track by demonstrating that a hybrid end-to-end Diarizen framework with WavLM embeddings and optimized agglomerative hierarchical clustering significantly outperformed a SpeechBrain baseline, reducing the diarization error rate to 9.21% on the evaluation set.
This paper presents a computationally efficient, detection-gated deep learning pipeline that achieves state-of-the-art robustness and cross-dataset generalization in glottal segmentation from high-speed videoendoscopy, enabling reliable extraction of clinical biomarkers for distinguishing healthy from pathological vocal function.
This paper proposes a robust framework combining the hybrid CoAtNet architecture with model soups ensembling to effectively classify Intangible Cultural Heritage images from the Mekong Delta, achieving state-of-the-art performance on the ICH-17 dataset by reducing variance and enhancing generalization in data-scarce, high-similarity settings.
This paper proposes and analyzes a personalized multi-agent average reward TD-learning algorithm that leverages joint linear approximation to learn shared subspaces and local heads, demonstrating convergence with linear speedup despite the challenges of environmental heterogeneity and Markovian sampling.
This paper proposes a hybrid neural network architecture that embeds an interpretable, -regularized vector autoregressive model within a convolutional autoencoder to effectively extract and visualize sparse temporal dynamics from two-photon calcium imaging data while preserving non-sparse spatial information.
This paper introduces GramCol and a motion-feature selection algorithm to generate Interpretable Motion-Attentive Maps (IMAPs) that effectively localize both motion and non-motion concepts in Video Diffusion Transformers without requiring gradient calculations or parameter updates.
This paper introduces CGL, a continual GUI learning framework that mitigates catastrophic forgetting by dynamically balancing Supervised Fine-Tuning and Reinforcement Learning through an entropy-guided proportion adjustment mechanism and a specialized gradient surgery strategy, validated by a new AndroidControl-CL benchmark.
This paper provides the first theoretical proof that Adam's second-moment normalization yields significantly sharper high-probability convergence guarantees ( dependence) compared to SGD ( dependence) under the classical bounded variance model, thereby explaining its empirical superiority.
This paper introduces Compositional Probe Decomposition (CPD) to demonstrate that linear disentanglement of geometric and compositional information in atomistic foundation models is primarily driven by task alignment rather than architecture, revealing a significant performance gradient where models trained on specific properties like HOMO-LUMO gaps outperform energy-trained models and exhibit symmetry-dependent information routing.
This paper introduces XInsight, a multi-agent framework that aligns psychological support with the Exploration-Insight-Action paradigm through a structured Reason-Intervene-Reflect cycle to enhance interpretability and therapeutic effectiveness in digital well-being applications, accompanied by the XInsight-Bench evaluation protocol.
This paper introduces vLLM Hook, an open-source plug-in that enables the programmable access and manipulation of internal model states within the vLLM inference engine to support advanced test-time alignment techniques such as adversarial prompt detection, enhanced RAG, and activation steering.