Vision-language models (VLMs) can locate an image region referred to by a text prompt and route the corresponding visual evidence to the output, yet the internal mechanism behind this behavior is not ...
While post-training improves the capabilities of large language models (LLMs), it generally lowers their output diversity and creativity, negatively impacting tasks that explicitly require creativity ...
Reward design remains a central bottleneck for autonomous robot policy improvement, especially in long-horizon manipulation tasks where sparse success labels provide too little signal and binary prefe...
In this paper, we consider Robbins' problem, which is a full information variant of the well-known secretary selection problem. In this version of the problem, the goal is to minimize the expected ran...
Penetration testing, the practice of simulating cyberattacks to identify vulnerabilities, is a complex sequential decision-making task that is inherently partially observable and features large action...
Free-form bones, that conform closely to the surface, can effectively capture non-rigid deformations, but lack a kinematic structure necessary for intuitive control. Thus, we propose a Scaffold-Skin R...
Diffusion large language models (dLLMs) offer parallel decoding and bidirectional context, but state-of-the-art dLLMs require billions of parameters for competitive performance. While existing distill...
The mesopelagic zone, between 100 and 1000 m depth, contains up to 90% of marine vertebrate biomass. Mesopelagic fishes are key components of these ecosystems through their feeding interactions and pl...
Classical bond percolation theory studies the conditions for a given point in a random graph to be connected to infinity, or "escape" to infinity, via a sequence of random edges. In this work, we pres...
The deep ocean hosts autochthonous pressure-adapted microorganisms that are unique to this environment, as well as allochthonous pressure-sensitive members transported from shallow depths by vertical ...
Field observations and limited experimental studies indicate that elasmobranchs can repair substantial skin injuries, but the temporal course and cellular composition of wound healing in Atlantic spin...
Deep neural networks often exploit spurious associations in their training data, a failure known as shortcut learning. Concept-based explainability methods screen for shortcuts by testing whether conc...
Digital advertising increasingly relies on visual content, yet marketers lack rigorous methods for understanding how specific visual attributes causally affect consumer engagement. This paper addresse...
We develop Structured-Knowledge-Informed Neural Networks (SKINNs), a unified estimation framework that embeds theoretical, simulated, previously learned, or cross-domain insights as differentiable con...
In modern applications of linear mixed models, the number of candidate fixed-effects covariates can grow exponentially with the sample size, while dependence induced by random effects and possible dat...
Standard masked discrete diffusion models face limitations in reasoning tasks due to their inability to correct their own mistakes on the masking path. Since they rely on a fixed number of denoising s...
Underwater tactile force sensing is crucial for achieving nondestructive and stable object manipulation in ocean robotics, especially in deep-sea environments where traditional vision-based methods ar...
We study outlier tokens in Diffusion Transformers (DiTs) for image generation. Prior work has shown that Vision Transformers (ViTs) can produce a small number of high-norm tokens that attract dispropo...
We study minimum-norm interpolation (MNI) in overparameterized linear regression with isotropic Gaussian covariates, in settings where the MNI has no closed-form formula. Whereas most prior work relie...
🌊
