10569 papers
This paper introduces FSMLP, a novel time series forecasting framework that employs a Simplex-MLP layer with weights constrained to a standard simplex to theoretically reduce overfitting in channel-wise dependencies via Rademacher complexity analysis, thereby achieving superior accuracy and scalability across multiple benchmarks.
This paper addresses the lack of unified memory evaluation in Reinforcement Learning by introducing cognitive science-inspired definitions to classify agent memory types and proposing a standardized experimental methodology to ensure accurate assessment and comparison of memory capabilities.
The paper proposes LoCo-RLHF, a framework that leverages low-rank contextual modeling and a pessimistic reduced-subspace policy to effectively align large language models with heterogeneous human feedback while ensuring computational efficiency and robustness to distributional shifts.
This paper theoretically demonstrates and empirically validates that removing difficult examples, along with techniques like margin tuning and temperature scaling, enhances the generalization and downstream performance of unsupervised contrastive learning by mitigating the negative impact these examples have on the model's learning mechanism.
This paper identifies and empirically validates "preference leakage," a pervasive contamination problem where LLM judges exhibit significant bias toward student models with which they share a relationship (such as being the same model, having an inheritance link, or belonging to the same family), thereby challenging the reliability of LLM-as-a-judge evaluation paradigms.
This paper introduces MIKASA, a comprehensive benchmark suite featuring a classification framework and two distinct datasets (MIKASA-Base and MIKASA-Robo) to systematically evaluate and advance memory-enhanced reinforcement learning agents across diverse scenarios, with a specific focus on tabletop robotic manipulation.
This paper introduces the UNB StepUP-P150 dataset, a large-scale, high-resolution collection of plantar pressure data from 150 individuals across varying walking speeds and footwear conditions, designed to advance research in biometric gait recognition, biomechanics, and deep learning.
This paper introduces Implicit U-KAN 2.0, a novel medical image segmentation model that combines second-order neural ordinary differential equations (SONO) with MultiKAN layers in a two-phase encoder-decoder architecture to achieve superior performance, enhanced interpretability, and dimension-independent approximation capabilities while reducing computational costs.
This paper proposes the TS4NAP approach, which leverages medical taxonomies (ICD-10-CM and ICD-10-PCS) and graph matching to enhance the accuracy and explainability of next-activity prediction in patient treatment planning using MIMIC-IV data.
This paper presents a large-scale analysis of 326 image classification models across nine quality dimensions beyond accuracy, revealing that vision-language models, self-supervised initialization, and dataset size significantly influence model behavior, and introduces the QUBA score to holistically rank and recommend models based on specific user needs.
This paper introduces \data, a dataset of fine-grained interaction prompts, and proposes \model, a novel framework leveraging Multimodal Large Language Models for prompt decomposition, image critique, and targeted refinement to significantly enhance the generation of complex object interactions in text-to-image synthesis.
This paper introduces PinRec, a novel outcome-conditioned, multi-token generative retrieval model developed for Pinterest that successfully balances performance, diversity, and efficiency to meet industrial-scale recommendation needs and multiple business metrics.
This paper introduces Noise-to-Meaning Recursive Self-Improvement (N2M-RSI), a minimal formal model demonstrating that AI agents feeding their own outputs back as inputs can trigger unbounded internal complexity growth once a specific information-integration threshold is crossed, while remaining implementation-agnostic and scalable to agent swarms.
This paper introduces Akkumula, a scalable and transparent framework that utilizes Spiking Neural Networks to model realistic driver behavior through evidence accumulation, effectively reproducing braking, accelerating, and steering actions while overcoming the limitations of existing hand-crafted approaches.
This paper introduces PubHealthBench, a new benchmark comprising over 8,000 questions derived from UK government guidance, to evaluate LLMs on public health knowledge and finds that while state-of-the-art models excel in multiple-choice tasks, their performance on free-form responses remains limited, highlighting the need for additional safeguards in real-world applications.
This paper establishes the theoretical foundations and single-agent architecture of the Emotion-Gradient Metacognitive RSI (EG-MRSI) framework, a novel system that integrates introspective metacognition and emotion-driven intrinsic motivation to enable provably safe, recursive self-improvement through differentiable reward signals and quantifiable semantic learning metrics.
This paper proposes a novel unsupervised framework for Invariant Risk Minimization that redefines invariance through feature distribution alignment, introducing the linear PICA and deep generative VIAE methods to learn robust, environment-invariant representations from unlabeled data.
TSPulse is a family of ultra-lightweight, pre-trained time-series models that utilize a novel disentanglement framework to learn complementary temporal, spectral, and semantic views, achieving state-of-the-art zero-shot and fine-tuned performance across diverse diagnostic tasks while outperforming models 10–100 times larger.
This paper establishes a tighter information-theoretic lower bound and proposes a modified Track-and-Stop algorithm with a tie-aware stopping rule that achieves asymptotic instance-optimality for best-arm identification in stochastic multi-armed bandits when the number of optimal arms is known.
This paper proposes "Feature Mixing," an extremely simple and fast modality-agnostic method for multimodal outlier synthesis that achieves state-of-the-art performance in out-of-distribution detection and segmentation while offering significant speedups, alongside the introduction of a new multimodal dataset called CARLA-OOD.