Hasil untuk "Law"

H. Jeffreys

It is shown that a certain differential form depending on the values of the parameters in a law of chance is invariant for all transformations of the parameters when the law is differentiable with regard to all parameters. For laws containing a location and a scale parameter a form with a somewhat restricted type of invariance is found even when the law is not everywhere differentiable with regard to the parameters. This form has the properties required to give a general rule for stating the prior probability in a large class of estimation problems.

C. Borgnakke, R. Sonntag

Law. Policy

Danny Hernandez, J. Kaplan, T. Henighan et al.

We study empirical scaling laws for transfer learning between distributions in an unsupervised, fine-tuning setting. When we train increasingly large neural networks from-scratch on a fixed-size dataset, they eventually become data-limited and stop improving in performance (cross-entropy loss). When we do the same for models pre-trained on a large language dataset, the slope in performance gains is merely reduced rather than going to zero. We calculate the effective data"transferred"from pre-training by determining how much data a transformer of the same size would have required to achieve the same loss when training from scratch. In other words, we focus on units of data while holding everything else fixed. We find that the effective data transferred is described well in the low data regime by a power-law of parameter count and fine-tuning dataset size. We believe the exponents in these power-laws correspond to measures of the generality of a model and proximity of distributions (in a directed rather than symmetric sense). We find that pre-training effectively multiplies the fine-tuning dataset size. Transfer, like overall performance, scales predictably in terms of parameters, data, and compute.

Alberto Abad Suárez Ávila

El texto aborda la prohibición expresa para decretar la suspensión con efectos generales en los medios de control de constitucionalidad de normas generales a cargo de los órganos jurisdiccionales del Poder Judicial de la Federación, conforme a las modificaciones a los artículos 105 y 107 de la Constitución mexicana del 15 de septiembre de 2024. ¿Cuál ha sido la trayectoria de la figura dentro de la jurisdicción constitucional mexicana? ¿La reforma acerca o aleja a la jurisdicción constitucional mexicana de las mejores prácticas en el derecho comparado? El objetivo es responder a estos cuestionamientos a través de un análisis de la trayectoria histórica de la figura en el derecho constitucional mexicano, con referencia a un bench-marking de las mejores prácticas en el derecho comparado.

Qihua Zhang, Christopher J. Jensen, Alexander J. Grutter et al.

Altermagnets are a recently-discovered class of materials with magnetic ordering that have a zero net magnetization and a momentum-dependent spin splitting in their band structure, arising from a collinear spin arrangement with alternating polarizations in the crystal lattice. The nickeline-structured manganese telluride (α-MnTe) is an attractive altermagnet candidate due to its predicted large spin splitting energy and a transition temperature near 300K. In this work, we present a thorough investigation of the spin structure of α-MnTe thin films grown by molecular beam epitaxy with very high crystal quality and low residual magnetization. The epitaxial α-MnTe films have a full-width-at-half-maximum of 0.1° as measured by x-ray-diffraction rocking curves and a root-mean-square roughness below 1 nm. Neutron diffraction measurements confirm the antiferromagnetic order in the α-MnTe film and show a Néel temperature of 307 K. Polarized neutron reflectometry detects a vanishingly small net magnetization which may be confined to the MnTe/InP interface, highlighting the near-ideal stoichiometry in the sample. In vacuo angle resolved photoemission spectroscopy reveals that the bulk band spectrum of the MnTe films is consistent with the weak altermagnetic order as theoretically predicted and observed for the high symmetry nodal plane in the center of the Brillouin zone. This study establishes optimized growth conditions for the synthesis of stoichiometric α-MnTe thin films which exhibit exceptional structural and magnetic ordering, thereby providing a robust platform for the precise characterization of their altermagnetic properties.

Cristina-Matilda VĂNOAGĂ

The present paper presents some of the latest Romanian legal provisions to support the European Green Deal objectives and the adoption of the "Fit for 55" package. Our paper presents some of the latest Romanian legislative initiatives and programs as part of the joint European effort to make Europe the first green continent. Thus, we briefly discuss The National Integrated Energy and Climate Change Plan 2021-2030, Emergency Ordinance 71/2021 on the promotion of non-polluting road transport vehicles, in support of low-emission mobility, Rabla Plus Program 2024, Emergency Ordinance 108/2022 regarding the decarbonisation of the energy sector, the proposed National Hydrogen Strategy and presents the modifications to the Law 220/ 2008.

Oren Neumann, Claudius Gros

Neural scaling laws are observed in a range of domains, to date with no universal understanding of why they occur. Recent theories suggest that loss power laws arise from Zipf's law, a power law observed in domains like natural language. One theory suggests that language scaling laws emerge when Zipf-distributed task quanta are learned in descending order of frequency. In this paper we examine power-law scaling in AlphaZero, a reinforcement learning algorithm, using a model of language-model scaling. We find that game states in training and inference data scale with Zipf's law, which is known to arise from the tree structure of the environment, and examine the correlation between scaling-law and Zipf's-law exponents. In agreement with the quanta scaling model, we find that agents optimize state loss in descending order of frequency, even though this order scales inversely with modelling complexity. We also find that inverse scaling, the failure of models to improve with size, is correlated with unusual Zipf curves where end-game states are among the most frequent states. We show evidence that larger models shift their focus to these less-important states, sacrificing their understanding of important early-game states.

Lilian Edwards, Reuben Binns

Generative AI systems often generate outputs about real people, even when not explicitly prompted to do so. This can lead to significant reputational and privacy harms, especially when sensitive, misleading, and outright false. This paper considers what legal tools currently exist to protect such individuals, with a particular focus on defamation and data protection law. We explore the potential of libel law, arguing that it is a potential but not an ideal remedy, due to lack of harmonization, and the focus on damages rather than systematic prevention of future libel. We then turn to data protection law, arguing that the data subject rights to erasure and rectification may offer some more meaningful protection, although the technical feasibility of compliance is a matter of ongoing research. We conclude by noting the limitations of these individualistic remedies and hint at the need for a more systemic, environmental approach to protecting the infosphere against generative AI.

Anastasis Kratsios, Haitz Sáez de Ocáriz Borde, Takashi Furuya et al.

Mixture-of-Experts (MoEs) can scale up beyond traditional deep learning models by employing a routing strategy in which each input is processed by a single "expert" deep learning model. This strategy allows us to scale up the number of parameters defining the MoE while maintaining sparse activation, i.e., MoEs only load a small number of their total parameters into GPU VRAM for the forward pass depending on the input. In this paper, we provide an approximation and learning-theoretic analysis of mixtures of expert MLPs with (P)ReLU activation functions. We first prove that for every error level $\varepsilon>0$ and every Lipschitz function $f:[0,1]^n\to \mathbb{R}$, one can construct a MoMLP model (a Mixture-of-Experts comprising of (P)ReLU MLPs) which uniformly approximates $f$ to $\varepsilon$ accuracy over $[0,1]^n$, while only requiring networks of $\mathcal{O}(\varepsilon^{-1})$ parameters to be loaded in memory. Additionally, we show that MoMLPs can generalize since the entire MoMLP model has a (finite) VC dimension of $\tilde{O}(L\max\{nL,JW\})$, if there are $L$ experts and each expert has a depth and width of $J$ and $W$, respectively.

El-Mahdi El-Mhamdi, Lê-Nguyên Hoang

``When a measure becomes a target, it ceases to be a good measure'', this adage is known as {\it Goodhart's law}. In this paper, we investigate formally this law and prove that it critically depends on the tail distribution of the discrepancy between the true goal and the measure that is optimized. Discrepancies with long-tail distributions favor a Goodhart's law, that is, the optimization of the measure can have a counter-productive effect on the goal. We provide a formal setting to assess Goodhart's law by studying the asymptotic behavior of the correlation between the goal and the measure, as the measure is optimized. Moreover, we introduce a distinction between a {\it weak} Goodhart's law, when over-optimizing the metric is useless for the true goal, and a {\it strong} Goodhart's law, when over-optimizing the metric is harmful for the true goal. A distinction which we prove to depend on the tail distribution. We stress the implications of this result to large-scale decision making and policies that are (and have to be) based on metrics, and propose numerous research directions to better assess the safety of such policies in general, and to the particularly concerning case where these policies are automated with algorithms.

Vindria Shafa Clarissha, Recca Ayu Hapsari, Yulia Hesti

Mortgage rights are regulated in Law Number four of 1996 concerning Mortgage Rights on Land and Objects Related to Land, hereinafter abbreviated as UUHT. With a variety of options 13/pdt.bth/2021/pn/tjk where the land due to the marriage is used as collateral for the rights of borrowing and borrowing through the husband and spouse as guarantors. election No.13/pdt.bth/2021/pn/tjk. The approach used in this study is empirical through interviews with judges at the Tanjung Karang District Court. Land rights that can be encumbered with Borrowing Rights consist of: Ownership Rights, Cultivation Rights, Building Use Rights, and Land Use Rights on state land which in accordance with the legal guidelines and guidelines of the winner must be registered and according to their nature can be transferred. Flats that stand on Ownership Rights, Building Use Rights and Use Rights granted through the state. Because one of the parties,namely Edwin Bunyamin Pohar,was not drawn as a party in the a quo rebuttal case,the rebuttal of the rebuttal as a rebuttal is lacking because not all parties in the main case are parties in the a quo rebuttal case, so the rebuttal of the rebuttal must be declared no acceptable (Niet Onvankelijke Verklaard).

Kang Niu, Xi Chen, Di Yang et al.

To intercept the great maneuvering target, combining with the sliding mode and the extended disturbance observer, a new control algorithm for integrated guidance and control (IGC) system is proposed in this paper. Firstly, the paper formulates the Missile–Target problem. Then the paper establishes an uncertain IGC dynamic model where the nonlinearities, the perturbations and the maneuvering of the target are regarded as disturbance. Secondly, a second-order disturbance observer is designed to estimate the disturbance and their derivatives.. After this, combining with the second-order disturbance observer, a modified sliding surface and the corresponding reaching law are designed to obtain the rudder deflection command directly. Thus, the real sense of IGC system is achieved. Next, the paper uses the Lyapunov stability theory to prove the stability of the system. Finally, the paper provides different simulation cases, which have different maneuver modes of the target, to demonstrate the superiority of the proposed method in reducing the response time, increasing the rudder response, and having a high interception probability.

Paul H. Rubin

John E. Semonche, L. Friedman

This book is a general history of the legal system of the United States, beginning in the colonial period, and continuing up to the present. The work was originally published in 1973; this is the fourth edition, which brings the material up to date and incorporates recent research. The book covers the changing configurations of commercial law, criminal law, and family law, and the law of property; lays great stress on race relations, especially black-white relations; it deals also with the legal profession and legal education. The approach throughout is geared toward an intelligent lay audience. Legal jargon is avoided. The underlying theory of the book is that law is the product of society, so that what is attempted, in essence, is a more or less sociological history of the legal system, as it evolved over the years.

Rahul Vigneswaran, Marc T. Law, Vineeth N. Balasubramanian et al.

The visual world naturally exhibits an imbalance in the number of object or scene instances resulting in a \emph{long-tailed distribution}. This imbalance poses significant challenges for classification models based on deep learning. Oversampling instances of the tail classes attempts to solve this imbalance. However, the limited visual diversity results in a network with poor representation ability. A simple counter to this is decoupling the representation and classifier networks and using oversampling only to train the classifier. In this paper, instead of repeatedly re-sampling the same image (and thereby features), we explore a direction that attempts to generate meaningful features by estimating the tail category's distribution. Inspired by ideas from recent work on few-shot learning, we create calibrated distributions to sample additional features that are subsequently used to train the classifier. Through several experiments on the CIFAR-100-LT (long-tail) dataset with varying imbalance factors and on mini-ImageNet-LT (long-tail), we show the efficacy of our approach and establish a new state-of-the-art. We also present a qualitative analysis of generated features using t-SNE visualizations and analyze the nearest neighbors used to calibrate the tail class distributions. Our code is available at https://github.com/rahulvigneswaran/TailCalibX.

R. Herrnstein

G. Priest

Frederik Blicher Jepsen

Begić-Šarkinović Taida

This article addresses recent human rights matters in the context of treaty-based investment arbitration. After having analyzed two recent arbitral awards, Urbaser v. Argentina and Bear Creek v. Peru, the article argues that host states' counterclaims on the basis of human rights can serve as an effective vehicle for rebalancing investors' rights and regulatory powers of host states. Both cases illustrate a more progressive approach of investment arbitral tribunals when dealing with human rights issues. Host states' counterclaims can also provide a door for arbitrating corporate-related human rights violations in the sense that they can ensure the application of international human rights law in investment arbitration cases. International investment promotion for sustainable development requires rebalancing the rights and obligations of states and investors which could help ensure corporate human rights accountability. In this regard, international investment treaties have to be renegotiated in order to include more human rights content targeting investors' human rights conduct and in that way provide a more balanced approach when disputes arise.