Rare polar bear adoption could save cub's life The cubs were born into a well-studied'celebration' of polar bears in Canada. Breakthroughs, discoveries, and DIY tips sent every weekday. Scientists in Churchill, Manitoba, Canada (aka the polar bear capital of the world) have confirmed that a wild female polar bear has adopted a cub that is not her own. This rare behavior was captured on cameras during the polar bear's annual migration along Western Hudson Bay . Researchers from Environment and Climate Change Canada and Polar Bears International spotted the mother bear (designated as bear X33991) during spring 2025, when she came out of her maternity den.

Model Recovery (MR) enables safe, explainable decision making in mission-critical autonomous systems (MCAS) by learning governing dynamical equations, but its deployment on edge devices is hindered by the iterative nature of neural ordinary differential equations (NODEs), which are inefficient on FPGAs. Memory and energy consumption are the main concerns when applying MR on edge devices for real-time operation. We propose MERINDA, a novel FPGA-accelerated MR framework that replaces iterative solvers with a parallelizable neural architecture equivalent to NODEs. MERINDA achieves nearly 11x lower DRAM usage and 2.2x faster runtime compared to mobile GPUs. Experiments reveal an inverse relationship between memory and energy at fixed accuracy, highlighting MERINDA's suitability for resource-constrained, real-time MCAS.

We start by exploring strategies for eliciting explicit, coherent numerical predictive distributions from LLMs.

We will prove by the induction. Let's suppose that the formula holds for By the definition in Eq. 4 and the chain rule, we can get that: N In this section, we give error bounds for spline representation. In the present work, we focus on using spline for smoothing noisy data. Following [51], we have spline fitting error bounds, as following. Eq. 11 L Output: Mean estimation: θ A.4 Training Details Additional training hyper parameters used in Sec. 4 is shown in the Tab. 2. T able 2: Training Details We list additional discovery and UQ results in this section.

Automatic readability assessment plays a key role in ensuring effective and accessible written communication. Despite significant progress, the field is hindered by inconsistent definitions of readability and measurements that rely on surface-level text properties. In this work, we investigate the factors shaping human perceptions of readability through the analysis of 897 judgments, finding that, beyond surface-level cues, information content and topic strongly shape text comprehensibility. Furthermore, we evaluate 15 popular readability metrics across five English datasets, contrasting them with six more nuanced, model-based metrics. Our results show that four model-based metrics consistently place among the top four in rank correlations with human judgments, while the best performing traditional metric achieves an average rank of 8.6. These findings highlight a mismatch between current readability metrics and human perceptions, pointing to model-based approaches as a more promising direction.

We start by exploring strategies for eliciting explicit, coherent numerical predictive distributions from LLMs.

We will prove by the induction. Let's suppose that the formula holds for By the definition in Eq. 4 and the chain rule, we can get that: N In this section, we give error bounds for spline representation. In the present work, we focus on using spline for smoothing noisy data. Following [51], we have spline fitting error bounds, as following. Eq. 11 L Output: Mean estimation: θ A.4 Training Details Additional training hyper parameters used in Sec. 4 is shown in the Tab. 2. T able 2: Training Details We list additional discovery and UQ results in this section.

Large language models (LLMs) excel at various natural language processing tasks, but their tendency to generate hallucinations undermines their reliability. Existing hallucination detection methods leveraging hidden states predominantly focus on static and isolated representations, overlooking their dynamic evolution across layers, which limits efficacy. To address this limitation, we shift the focus to the hidden state update process and introduce a novel metric, the ICR Score (Information Contribution to Residual Stream), which quantifies the contribution of modules to the hidden states' update. We empirically validate that the ICR Score is effective and reliable in distinguishing hallucinations. Building on these insights, we propose a hallucination detection method, the ICR Probe, which captures the cross-layer evolution of hidden states. Experimental results show that the ICR Probe achieves superior performance with significantly fewer parameters. Furthermore, ablation studies and case analyses offer deeper insights into the underlying mechanism of this method, improving its interpretability.