搜索结果：Zhonghua nei ke za zhi

Evidence absence is not evidence insufficiency, but fact verification benchmarks can make them observationally similar. The Not Enough Information (NEI) label is often operationalized through different evidence conditions, and that choice silently determines what a verifier learns and what its score can hide. We introduce NEI-CAP, a construction-aware diagnostic protocol for insufficient-evidence evaluation. Each NEI example carries the construction family that produced it; NEI-CAP audits shortcut cues, validates hard cases through human adjudication, and tests whether competence transfers across constructions. We instantiate the protocol in SciFact-style scientific verification, with FEVER and HoVer as bounded external controls. Across these settings, NEI competence does not transfer reliably: models trained on shortcut-prone constructions fail to recognize semantically related insufficient evidence, and mixed-construction training narrows but does not close the gap. Fixed-claim diagnostics further show that the evidence condition shifts confidence in the reference Support/Refute label, not only NEI recall, so an aggregate NEI score can hide which problem a model has actually solv

Existing humor datasets and evaluations predominantly focus on English, lacking resources for culturally nuanced humor in non-English languages like Chinese. To address this gap, we construct Chumor, a dataset sourced from Ruo Zhi Ba (RZB), a Chinese Reddit-like platform dedicated to sharing intellectually challenging and culturally specific jokes. We annotate explanations for each joke and evaluate human explanations against two state-of-the-art LLMs, GPT-4o and ERNIE Bot, through A/B testing by native Chinese speakers. Our evaluation shows that Chumor is challenging even for SOTA LLMs, and the human explanations for Chumor jokes are significantly better than explanations generated by the LLMs.

A dynamic light scattering study of director-layer fluctuations in the antiferroelectric smectic-ZA phase of the ferroelectric nematic liquid crystal DIO is reported. The dynamics are consistent with the distinctive feature of the ZA phase that the smectic layers form parallel to the axis of molecular orientational order (director). A model is developed to describe quantitatively the dispersion of the fluctuation relaxation rates. The model is based on a specialization of the elastic free energy density of the smectic-C phase to the case of 90 degree director tilt, a "first-order" approximation of the viscous stresses by their form for an incompressible uniaxial fluid, and a treatment of the effect of chevron layer structure that develops in planar sample cells due to temperature-dependent layer shrinkage, as documented in previous studies on DIO. From the modeling, the layer compression elastic constant is estimated to be ~100 times lower in the smectic-ZA phase than in an ordinary smectic-A liquid crystal. Possible effects of the antiferroelectric layer polarization on the director splay elasticity and viscosity are described. The temperature dependencies of the splay, twist, and

Let $\mathsf{mod} R$ denote the category of finitely generated $R$-modules for a commutative noetherian ring $R$. In this paper, we investigate KE-closed subcategories of $\mathsf{mod} R$ as a continuation of our previous work. We associate a function on $\mathrm{Spec} R$ with each KE-closed subcategory of $\mathsf{mod} R$, and show that this function completely determines the original subcategory. To classify the functions obtained from KE-closed subcategories, we introduce the notion of an $n$-Bass function for each $n\ge 0$. We obtain a bijection between the set of KE-closed subcategories and the set of $2$-Bass functions provided that $R$ is $(S_2)$-excellent in the sense of Česnavičius.

Let $R$ be a commutative noetherian ring and denote by $\mathsf{mod} R$ the category of finitely generated $R$-modules. In this paper, we study KE-closed subcategories of $\mathsf{mod} R$, that is, additive subcategories closed under kernels and extensions. We first give a characterization of KE-closed subcategories: a KE-closed subcategory is a torsion-free class in a torsion-free class. As an immediate application of the dual statement, we give a conceptual proof of Stanley-Wang's result about narrow subcategories. Next, we classify the KE-closed subcategories of $\mathsf{mod} R$ when $\mathrm{dim} R \le 1$ and when $R$ is a two-dimensional normal domain. More precisely, in the former case, we prove that KE-closed subcategories coincide with torsion-free classes in $\mathsf{mod} R$. Moreover, this condition implies $\mathrm{dim} R \le 1$ when $R$ is a homomorphic image of a Cohen-Macaulay ring (e.g. a finitely generated algebra over a regular ring). Thus, we give a complete answer for the title.

A relevant collection is a collection, $F$, of sets, such that each set in $F$ has the same cardinality, $α(F)$. A Konig Egervary (KE) collection is a relevant collection $F$, that satisfies $|\bigcup F|+|\bigcap F|=2α(F)$. An hke (hereditary KE) collection is a relevant collection such that all of his non-empty subsets are KE collections. In \cite{jlm} and \cite{dam}, Jarden, Levit and Mandrescu presented results concerning graphs, that give the motivation for the study of hke collections. In \cite{hke}, Jarden characterize hke collections. Let $Γ$ be a relevant collection such that $Γ-\{S\}$ is an hke collection, for every $S \in Γ$. We study the difference between $|\bigcap Γ_1-\bigcup Γ_2|$ and $|\bigcap Γ_2-\bigcup Γ_1|$, where $\{Γ_1,Γ_2\}$ is a partition of $Γ$. We get new characterizations for an hke collection and for a KE graph.

The supernova remnant (SNR) Kes 75/PSR J1846-0258 association can be regarded as certain due to the accurate location of young PSR J1846-0258 at the center of Kes 75 and the detected bright radio/X-ray synchrotron nebula surrounding the pulsar. We provide a new distance estimate to the SNR/pulsar system by analyzing the HI and $^{13}$CO maps, the HI emission and absorption spectra, and the $^{13}$CO emission spectrum of Kes 75. No absorption features at negative velocities strongly argue against the widely-used large distance of 19 to 21 kpc for Kes 75, and show that Kes 75 is within the Solar circle, i.e. a distance $d<$13.2 kpc. Kes 75 is likely at distance of 5.1 to 7.5 kpc because the highest HI absorption velocity is at 95 km/s and no absorption is associated with a nearby HI emission peak at 102 km/s in the direction of Kes 75. This distance to Kes 75 gives a reasonable luminosity of PSR J1846-0258 and its PWN, and also leads to a much smaller radius for Kes 75. So the age of the SNR is consistent with the spin-down age of PSR J1846-0258, confirming this pulsar as the second-youngest in the Galaxy.

A common ingredient in cosmological perturbation theory (PT) is the expansion of the dark matter overdensity $δ$ in the Lagrangian displacement $s$, which amounts to enforcing mass conservation perturbatively. In Eulerian PT (EPT), that expansion occurs already at the level of the continuity equation; in Lagrangian PT (LPT) it is done in the Poisson equation. We show that the resulting perturbative solutions for $δ$ can diverge not because of the expansion in $s$ per se, but because of an exchange of an infinite sum with a Fourier integral that violates the conditions of Lebesgue's dominated-convergence (DC) theorem. We show that this DC obstruction (DCO) is one clear reason why the convergence of EPT is controlled by advection terms beyond the linear $δ$. The same DCO underlies LPT: LPT's region of validity is the resummation region of a DC-violating series, bounded by shell crossing on one side and severely underdense regions on the other. Effective field theories (EFT) of large-scale structure need to smooth at short scales just to recover from that DCO, independent of whether non-linearities beyond mass conservation are important or not. An alternative is to never expand $δ$ in

We provide a new distance estimate to the supernova remnant (SNR) Kes 73 and its associated anomalous X-ray pulsar (AXP) 1E 1841-045. 21 cm HI images and HI absorption/ emission spectra from new VLA observations, and 13CO emission spectra of Kes 73 and two adjacent compact HII regions (G27.276+0.148 and G27.491+0.189) are analyzed. The HI images show prominent absorption features associated with Kes 73 and the HII regions. The absorption appears up to the tangent point velocity giving a lower distance limit to Kes 73 of 7.5 kpc, which has previously been given as the upper limit. Also, G27.276+0.148 and G27.491+0.189 are at the far kinematic distances of their radio recombination line velocities. There is prominent HI emission in the range 80--90 km/s for all three objects. The two HII regions show HI absorption at ~ 84 km/s, but there is no absorption in the Kes 73 absorption spectrum. This implies an upper distance limit of ~ 9.8 kpc to Kes 73. This corrected larger distance to Kes 73/ AXP 1E 1841-045 system leads to a refined age of the SNR of 500 to 1000 yr, and a ~ 50% larger AXP X-ray luminosity.

We study the structure and evolution of dark matter halos from z = 300 to z = 6 for two cosmological N-body simulation initialization techniques. While the second order Lagrangian perturbation theory (2LPT) and the Zel'dovich approximation (ZA) both produce accurate present day halo mass functions, earlier collapse of dense regions in 2LPT can result in larger mass halos at high redshift. We explore the differences in dark matter halo mass and concentration due to initialization method through three 2LPT and three ZA initialized cosmological simulations. We find that 2LPT induces more rapid halo growth, resulting in more massive halos compared to ZA. This effect is most pronounced for high mass halos and at high redshift. Halo concentration is, on average, largely similar between 2LPT and ZA, but retains differences when viewed as a function of halo mass. For both mass and concentration, the difference between typical individual halos can be very large, highlighting the shortcomings of ZA-initialized simulations for high-z halo population studies.

Knowledge graphs have emerged as a key abstraction for organizing information in diverse domains and their embeddings are increasingly used to harness their information in various information retrieval and machine learning tasks. However, the ever growing size of knowledge graphs requires computationally efficient algorithms capable of scaling to graphs with millions of nodes and billions of edges. This paper presents DGL-KE, an open-source package to efficiently compute knowledge graph embeddings. DGL-KE introduces various novel optimizations that accelerate training on knowledge graphs with millions of nodes and billions of edges using multi-processing, multi-GPU, and distributed parallelism. These optimizations are designed to increase data locality, reduce communication overhead, overlap computations with memory accesses, and achieve high operation efficiency. Experiments on knowledge graphs consisting of over 86M nodes and 338M edges show that DGL-KE can compute embeddings in 100 minutes on an EC2 instance with 8 GPUs and 30 minutes on an EC2 cluster with 4 machines with 48 cores/machine. These results represent a 2x~5x speedup over the best competing approaches. DGL-KE is ava

I report here on the analysis and interpretation of a Chandra observation of the supernova remnant Kes 32. Kes 32 is rather weak in X-rays due to a large interstellar absorption, which is found to be ~4E22 cm^-2, larger than previously reported. Spectral analysis indicates that the ionization age of this object is very young, with n_e t ~ 4E9 cm^-3s, and a temperature of kT_e ~ 1 keV. The X-ray emission peaks at a smaller radius than in the radio. The low ionization age suggests that Kes 32 is a young remnant. However, a young age is in contradiction with the relatively large apparent size, which indicates an age of several thousand years, instead of a few hundred years. This problem is discussed in connection with Kes 32's unknown distance and its possible association with the Norma galactic arm.

Based on the data from the \asca observation of SNRs Kes79 and W49B, we present here the analysis of their X-ray spectra and morphologies. The Kes79 spectrum can be well fitted by a single NEI component, and the narrow-band images of that source show an inhomogeneous distribution of heavy elements. The heavy elements are richest in the positions S, SE and SW of Kes79, where there may exist interaction between shocks and molecular clouds implied by radio observations. For W49B we present here the non-equilibrium ionization (NEI) analysis based on its emission line diagnostics, and the spectral fit using two NEI components. The reverse shock in W49B may be still hot and we don't find evidence for a hotter blast wave in \asca spectra.

Multi-hop Question Answering (MQA) under knowledge editing (KE) is a key challenge in Large Language Models (LLMs). While best-performing solutions in this domain use a plan and solve paradigm to split a question into sub-questions followed by response generation, we claim that this approach is sub-optimal as it fails for hard to decompose questions, and it does not explicitly cater to correlated knowledge updates resulting as a consequence of knowledge edits. This has a detrimental impact on the overall consistency of the updated knowledge. To address these issues, in this paper, we propose a novel framework named RULE-KE, i.e., RULE based Knowledge Editing, which is a cherry on the top for augmenting the performance of all existing MQA methods under KE. Specifically, RULE-KE leverages rule discovery to discover a set of logical rules. Then, it uses these discovered rules to update knowledge about facts highly correlated with the edit. Experimental evaluation using existing and newly curated datasets (i.e., RKE-EVAL) shows that RULE-KE helps augment both performances of parameter-based and memory-based solutions up to 92% and 112.9%, respectively.

A 30 ks \chandra ACIS-I observation of Kes 79 reveals rich spatial structures, including many filaments, three partial shells, a loop and a ``protrusion''. Most of them have corresponding radio features. Regardless of the different results from two non-equilibrium ionization (NEI) codes, temperatures of different parts of the remnant are all around 0.7 keV, which is surprisingly constant for a remnant with such rich structure. If thermal conduction is responsible for smoothing the temperature gradient, a lower limit on the thermal conductivity of $\sim$ 1/10 of the Spitzer value can be derived. Thus, thermal conduction may play an important role in the evolution of at least some SNRs. No spectral signature of the ejecta is found, which suggests the ejecta material has been well mixed with the ambient medium. From the morphology and the spectral properties, we suggest the bright inner shell is a wind-driven shell (WDS) overtaken by the blast wave (the outer shell) and estimate the age of the remnant to be $\sim$ 6 kyr for the assumed dynamics. Projection is also required to explain the complicated morphology of Kes 79.

Chain-of-thought (CoT) monitoring is a promising safety mechanism for AI agents, based on the premise that visible reasoning traces can surface misaligned or deceptive behavior. While effective in standard scenarios, recent work highlights that LLMs remain vulnerable to persuasion-based jailbreaks, where natural-language arguments override model constraints. We stress-test whether this vulnerability extends to monitoring LLMs: can an adversarial agent persuade its CoT monitor to approve proposed actions that violate the monitor's policy? We design an evaluation framework with 40 tasks and analyze thousands of agent-monitor interactions, where agents are instructed to argue for policy-violating proposals. We find that in such adversarial settings, monitor access to the agent's CoT reasoning increases rather than decreases approval of harmful actions on average by 9.5%, as the scratchpad provides an additional persuasion channel. To address this, we introduce a fact-checking monitoring framework. We find that a fact-checker and monitor pairing from different model families, for example a Claude 3.7 Sonnet monitor paired with a GPT-4.1 fact-checker, reduces approval of policy-violatin

We describe an improved nonequilibrium ionization (NEI) method that we have developed as an optional module for the FLASH magnetohydrodynamic simulation code. The method employs an eigenvalue approach rather than the earlier iterative ordinary differential equation approach to solve the stiff differential equations involved in NEI calculations. The new code also allows the atomic data to be easily updated from the AtomDB database. We compare both the updated atomic data and the methods separately. The new atomic data are shown to make a significant difference in some circumstances, although the general trends remain the same. Additionally, the new method also allows simultaneous calculation of the nonequilibrium radiative cooling, which is not included in the original method. The eigenvalue method improves the calculation efficiency overall with no loss of accuracy. We explore some common ways to present the NEI state with a sample simulation and find that using the average ionic charge difference from the equilibrium tends to be the clearest method.

The GPT-4o represents a significant milestone in enabling real-time interaction with large language models (LLMs) through speech, its remarkable low latency and high fluency not only capture attention but also stimulate research interest in the field. This real-time speech interaction is particularly valuable in scenarios requiring rapid feedback and immediate responses, dramatically enhancing user experience. However, there is a notable lack of research focused on real-time large speech language models, particularly for Chinese. In this work, we present KE-Omni, a seamless large speech language model built upon Ke-SpeechChat, a large-scale high-quality synthetic speech interaction dataset consisting of 7 million Chinese and English conversations, featuring 42,002 speakers, and totaling over 60,000 hours, This contributes significantly to the advancement of research and development in this field. The demos can be accessed at \url{https://huggingface.co/spaces/KE-Team/KE-Omni}.

Weakly-supervised point cloud segmentation with extremely limited labels is highly desirable to alleviate the expensive costs of collecting densely annotated 3D points. This paper explores applying the consistency regularization that is commonly used in weakly-supervised learning, for its point cloud counterpart with multiple data-specific augmentations, which has not been well studied. We observe that the straightforward way of applying consistency constraints to weakly-supervised point cloud segmentation has two major limitations: noisy pseudo labels due to the conventional confidence-based selection and insufficient consistency constraints due to discarding unreliable pseudo labels. Therefore, we propose a novel Reliability-Adaptive Consistency Network (RAC-Net) to use both prediction confidence and model uncertainty to measure the reliability of pseudo labels and apply consistency training on all unlabeled points while with different consistency constraints for different points based on the reliability of corresponding pseudo labels. Experimental results on the S3DIS and ScanNet-v2 benchmark datasets show that our model achieves superior performance in weakly-supervised point c