5406 papers
This paper proposes Soft Equivariance Regularization (SER), a lightweight, plug-in method that decouples invariance and equivariance objectives by enforcing equivariance on intermediate spatial features while preserving invariance on the final embedding, thereby improving both linear evaluation accuracy and robustness to geometric perturbations without requiring auxiliary heads or transformation labels.
This paper argues that RGB-only Multimodal Large Language Models fail to generalize across different cameras due to entangled perspective and object properties, and proposes a Camera-Aware MLLM framework that integrates camera intrinsics, augmented data, and 3D geometric priors to achieve robust, generalizable spatial intelligence.
This paper enhances the reliability of solar flare regression by applying conformal prediction to deep learning models, demonstrating that conformalized quantile regression outperforms alternative methods in achieving valid coverage rates and favorable interval lengths for space weather forecasting.
The paper introduces ATLAS, a reinforcement finetuning framework that enables small language models to effectively navigate large toolspaces by learning adaptive context acquisition and execution strategies, thereby achieving frontier-level performance with significantly reduced parameter and context budgets.
This paper presents an integrated pipeline combining knowledge-grounded generative modeling with automated LLM-based auditing to produce clinically consistent, privacy-preserving synthetic patient trajectories that overcome the limitations of existing methods by eliminating clinical inconsistencies while maintaining high statistical fidelity and downstream utility.
The paper introduces ProtAlign, a contrastive learning framework that unifies protein sequence and structure representations into a shared embedding space, thereby enabling cross-modal retrieval and improving downstream tasks like function annotation and stability estimation.
This paper proposes a bi-directional feedback fusion framework that integrates human activity embeddings with dual-timescale temporal modules to significantly improve the accuracy and interpretability of indoor CO2 and PM2.5 forecasting compared to traditional data-driven models.
This paper introduces FutureBoosting, a hybrid-AI framework that enhances electricity price forecasting by integrating forecasted features from a frozen time series foundation model into a regression model, thereby achieving significant accuracy improvements over state-of-the-art baselines while maintaining interpretability.
The paper proposes Safe Transformer, a modular approach that inserts an explicit, interpretable safety bit into pre-trained language models to achieve controllable alignment and near-zero attack success rates through lightweight fine-tuning, addressing the opacity of traditional implicit safety methods.
This paper introduces Orion, the first open end-to-end system that bypasses Apple's opaque CoreML framework to enable direct Neural Engine programming for large language model training and inference, achieving an 8.5x speedup in weight updates through a novel patching mechanism and demonstrating stable training of 110M-parameter models on Apple Silicon.
This paper proposes a reinforcement learning approach for dense crowd navigation that achieves zero-shot generalization to higher crowd densities by combining density-invariant observation encoding, density-randomized training, and physics-informed proxemic reward shaping, thereby significantly outperforming existing learning-based and analytical methods in success rate and collision avoidance without freezing.
This paper presents PolyBlocks, a modular, MLIR-based compiler infrastructure that utilizes pass pipelines and analytical cost models to automatically generate high-performance code for AI chips, demonstrating competitive or superior performance against existing frameworks like Torch Inductor and XLA on NVIDIA GPUs.
This paper introduces Calibrated Credit Intelligence (CCI), a deployment-oriented framework that integrates Bayesian uncertainty quantification, fairness-constrained gradient boosting, and shift-aware fusion to deliver accurate, reliable, and equitable credit risk scores that remain robust under temporal distribution shifts.
This paper proposes Rank-factorized Implicit Neural Bias (RIB), a novel positional bias mechanism that enables the use of hardware-efficient FlashAttention in Super-Resolution Transformers, allowing for significantly larger window sizes and training patches that achieve state-of-the-art performance (35.63 dB PSNR) while reducing training and inference times by 2.1 and 2.9, respectively.
This paper introduces an efficient framework for heterogeneous decentralized diffusion models that enables experts to train with mixed objectives (DDPM and Flow Matching) and reduced resource requirements, achieving a 16x decrease in compute and 14x reduction in data compared to prior approaches while improving image quality and diversity.
This paper proposes a framework that fine-tunes pretrained generative models to directly sample within complex, structured feasible regions, achieving a novel balance between strict constraint satisfaction and high-quality sample realism for safety-critical applications like robotics and autonomous driving.
This paper identifies that applying Group Relative Policy Optimization (GRPO) to diffusion language models causes reward collapse due to noisy importance ratio estimates and formulation mismatches, and proposes StableDRL, a reformulated algorithm featuring unconditional clipping and self-normalization to stabilize training and prevent policy drift.
The paper introduces DIRECTER, a novel activation steering method that dynamically modulates steering strength through a plausibility-guided decoding loop and layer sensitivity analysis to enhance LLM instruction-following accuracy while preventing the oversteering that typically degrades text quality.
This paper introduces ProtAlign, a multi-objective preference alignment framework that fine-tunes pretrained inverse folding models to simultaneously optimize diverse developability properties like solubility and thermostability while preserving structural designability, resulting in the enhanced MoMPNN model for practical protein sequence design.
This paper proposes the Latent Autoencoder Ensemble Kalman Filter (LAE-EnKF), a novel data assimilation method that learns a stable, linear state-space model in a latent space to overcome the performance limitations of standard EnKF on strongly nonlinear and chaotic systems while maintaining computational efficiency.