Related papers: Precise interpretations of traditional fine-tuning…
We consider highly inaccurate measurements made on classical stochastic and quantum systems. In the quantum case such a \e{weak} measurement preserves coherence between the system's alternatives. We demonstrate that in both cases the…
The Z line shape is measured at LEP with an accuracy at the per mille level. Usually it is described in the Standard Model of electroweak interactions with account of quantum corrections. Alternatively, one may attempt different…
Pretrained Foundation Models (PFMs) have transformed numerous applications by enabling efficient adaptation to customized tasks. Parameter-Efficient Fine-Tuning (PEFT) has emerged as a resource-efficient alternative to full fine-tuning,…
Foundation models pretrained on large-scale natural images are widely adapted to various cross-domain low-resource downstream tasks, benefiting from generalizable and transferable patterns captured by their representations. However, these…
Several reasonably model-independent formulations of the implications of new physics for precision electroweak measurements have been developed over the past years, most notably by Peskin and Takeuchi, and by Altarelli \etal. These…
The electron and muon magnetic moments have played, and continue to play, important roles in testing the fundamental mathematical description of physical reality called the Standard Model of particle physics (SM). The electron magnetic…
Supervised fine-tuning (SFT) is a standard approach to adapting large language models (LLMs) to new domains. In this work, we improve the statistical efficiency of SFT by selecting an informative subset of training examples. Specifically,…
An Effective Field Theory (EFT) analysis of the magnetic moment of the muon is discussed. The expression for the dipole moment is given in terms of operator coefficients of the low-energy effective field theory (LEFT) and the Standard Model…
Quantum filtering is a signal processing technique that estimates the posterior state of a quantum system under continuous measurements and has become a standard tool in quantum information processing, with applications in quantum state…
The implications of precision $Z$ and $W$-pole and neutral current data for testing the standard electroweak model, determining its parameters, and searching for new physics, are described.
Tabular Foundation Models (TFMs) have recently shown strong in-context learning capabilities on structured data, achieving zero-shot performance comparable to traditional machine learning methods. We find that zero-shot TFMs already achieve…
A brief review is made of some of the experimental signatures that may be associated to a certain class of extensions of the standard model. The material of these lectures is divided into two sections. After briefly sketching the present…
We have examined the electroweak radiative corrections in the LEP precision data in view of the new measurements of $M_W$ and $m_t$ as well as the recent progress in the higher order radiative corrections. From the minimal $\chi^2$-fit to…
We investigate the sensitivity of the FASER$\nu$ detector to new physics in the form of non-standard neutrino interactions. FASER$\nu$, which has recently been installed 480 m downstream of the ATLAS interaction point, will for the first…
Conventionally, supervised fine-tuning (SFT) is treated as a simple imitation learning process that only trains a policy to imitate expert behavior on demonstration datasets. In this work, we challenge this view by establishing a…
The potential of present and anticipated future electroweak precision data, including the Higgs boson and top quark masses, for testing quantum effects of the electroweak theory is investigated in the context of the Minimal Supersymmetric…
A novel approach to study electroweak physics at one-loop level in generic ${\rm SU(2)_L \times U(1)_Y}$ theories is introduced. It separates the 1-loop corrections into two pieces: process specific ones from vertex and box contributions,…
We review recent results that provide a new approach to the old problem of naturalness in supersymmetric models, without relying on subjective definitions for the fine-tuning associated with {\it fixing} the EW scale (to its measured value)…
While Supervised Fine-Tuning (SFT) and Rejection Sampling Fine-Tuning (RFT) are standard for LLM alignment, they either rely on costly expert data or discard valuable negative samples, leading to data inefficiency. To address this, we…
Packing, initially utilized in the pre-training phase, is an optimization technique designed to maximize hardware resource efficiency by combining different training sequences to fit the model's maximum input length. Although it has…