Alignment, interpretability, red-teaming, bias and privacy: the research on what these systems do when they misbehave.
arXiv:2607. 12332v1 Announce Type: new Abstract: We study the gradient flow dynamics of diagonal linear networks for regression tasks under infinitesimal initialization.
arXiv:2607. 12752v1 Announce Type: cross Abstract: While recent advances in 3D generation have enabled impressive visual synthesis, existing methods often rely on 2D diffusion supervision without explicit mechanisms for geometric consistency, leading to spatial hallucinations such as duplicated structures and misaligned geometry.
arXiv:2607. 12500v1 Announce Type: new Abstract: Deep learning models for online handwriting recognition have been shown effective and are increasingly deployed in practical applications.
arXiv:2607. 11919v1 Announce Type: cross Abstract: Human memory is reconstructive, not a faithful recording.
arXiv:2607. 12463v1 Announce Type: new Abstract: Coding agents must integrate external tool returns into ongoing reasoning - a capability that standard left-to-right pretraining on code exposes only in its forward direction.
arXiv:2607. 12474v1 Announce Type: new Abstract: Recent advances in foundation models have transformed AI for Science, enabling remarkably accurate predictive performance across domains ranging from protein folding to weather forecasting.
arXiv:2607. 12354v1 Announce Type: new Abstract: In this paper, we challenge the prevailing view that information dependency (including rote memorization) drives training data exposure to image reconstruction attacks.
arXiv:2607. 12255v1 Announce Type: new Abstract: We study interaction-aware mixture-of-experts for post-stroke rigidity prediction using multi-level views of structured health records.
arXiv:2604. 16232v2 Announce Type: replace-cross Abstract: Understanding natural and engineered systems often relies on symbolic formulations, such as differential equations, which provide interpretability and transferability beyond black-box models.
arXiv:2607. 12375v1 Announce Type: cross Abstract: Image Quality Assessment (IQA) in open-world environments remains challenging due to limited generalization and interpretability.
arXiv:2607. 11975v1 Announce Type: cross Abstract: Current machine unlearning methods predominantly rely on global, coarse-grained intervention strategies.
arXiv:2607. 11990v1 Announce Type: cross Abstract: Feedforward network (FFN) blocks account for a large fraction of the parameters and computation in Transformer architectures, yet their internal structure remains difficult to interpret due to the additive superposition induced by the residual stream.
arXiv:2607. 11890v1 Announce Type: cross Abstract: Open-ended surveys offer valuable insights, but they are notoriously difficult to analyze at scale.
arXiv:2607. 13007v1 Announce Type: new Abstract: Simulation-based algorithms are especially suited for high-uncertainty environments such as adversarial board games with significant elements of randomness and hidden information.
arXiv:2602. 15892v2 Announce Type: replace-cross Abstract: Visual perspective taking--inferring how the world appears from another's viewpoint--is foundational to social cognition.
arXiv:2607. 12631v1 Announce Type: cross Abstract: As Large Language Models (LLMs) are increasingly deployed as autonomous agents in high-stakes domains, understanding contextual factors that may modulate their decision-making becomes critical.
arXiv:2607. 11956v1 Announce Type: cross Abstract: Data Shapley is the standard principled answer to which training points are worth what, and its k-nearest-neighbor (KNN) specialization is the version deployed in practice: the exact estimator shipped by toolkits such as pyDVL and OpenDataVal.
arXiv:2512. 01241v4 Announce Type: replace-cross Abstract: Large language models (LLMs) and medical AI tools are routinely used by physicians and patients for medical advice, yet their clinical safety profiles remain poorly characterized.
arXiv:2512. 17788v2 Announce Type: replace Abstract: Multi-instance partial-label learning (MIPL) is a weakly supervised framework that extends the principles of multi-instance learning (MIL) and partial-label learning (PLL) to address the challenges of inexact supervision in both instance and label spaces.
arXiv:2607. 12281v1 Announce Type: cross Abstract: Transformer-style architectures are increasingly adopted for industrial recommendation systems, yet they inherit a design premise misaligned with the task: generative models rely on per-token autoregressive prediction, which justifies maintaining large intermediate tensors that scale with sequence length.