Related papers: Study of HERA data at Low Q^2 and Low x
The growing number of parameters and computational demands of large language models (LLMs) present significant challenges for their efficient deployment. Recently, there is an increasing interest in quantizing weights to extremely low…
A simple parametrisation of H1 and ZEUS data at HERA is given for the ranges in x and Q^2 of 10^{-4} - 5.10^{-2} and 5 - 250 GeV^2, respectively. This empirical expression is based on a strikingly similar dependence of the average charged…
We study of the accuracy of the Regge behavior of the gluon distribution function for obtain an approximation relation, which is frequently used to extract the logarithmic slopes of the structure function from the gluon distribution at…
We summarise existing experimental and theoretical knowledge on the structure function $F_2$ in the region of low $Q^2$ and low $x$. The constraints on the behaviour of structure functions in the limit $Q^2=0$ are listed. Phenomenological…
Possible interpretations of the HERA large-Q^2 data are reviewed briefly. The possibility of statistical fluctuations cannot be ruled out, and it seems premature to argue that the H1 and ZEUS anomalies are incompatible. The data cannot be…
Federated fine-tuning for Large Language Models (LLMs) faces significant challenges due to the heavy communication overhead of transmitting large model updates. Although Low Rank Adaptation (LoRA) has been proposed as a solution, yet its…
In the kinematic region of small $x$ and large $Q^2$ in deep inelastic scattering, presently being explored by HERA, we present an analysis of the evolution of the longitudinal structure function $F_L^{p}(x, Q^2)$ in the double scaling…
Low-Rank Adaptation (LoRA) is widely used for federated fine-tuning. Yet under non-IID settings, it can substantially underperform full-parameter fine-tuning. Through with-high-probability robustness analysis, we uncover that this gap can…
We compute the proton structure function F_2^p at small x and large Q^2 at next-to-leading order in alpha_s(Q^2), including summations of all leading and subleading logarithms of Q^2 and 1/x in a way consistent with momentum conservation.…
Fine-tuning adapts a pre-trained model to downstream tasks using a small amount of labeled data. Low-Rank Adaptation (LoRA) is an efficient fine-tuning method that reduces memory and computation costs while often achieving performance close…
We consider the behavior of the slope of the nuclear structure function $F_2^A$ in the kinematic region which will be explored in the eA HERA collider. We demonstrate that, similarly to the nucleon case, a `turn over' is predicted in this…
We provide an assessment of the impact of parton distributions on the determination of LHC processes, and of the accuracy with which parton distributions (PDFs) can be extracted from data, in particular from current and forthcoming HERA…
Recent results from the HERA ep collider are reviewed in these proceedings. The results are from measurements that probe QCD at high-energy scales, as defined by $Q^2$, the four-momentum-transfer squared of the collisions. These…
We propose a phenomenological study of the Balitsky-Fadin-Kuraev-Lipatov (BFKL) approach applied to the data on the proton structure function F_2 measured at HERA in the small-x_{Bjorken} region. In a first part we use a simplified…
The formulas directly connecting parton distribution functions (PDFs) at the leading (LO) and the next to leading (NLO) QCD orders are applied to both unpolarized and polarized valence PDFs and produce improved LO results on these…
A higher--twist nuclear enhancement of $R = \sigma_L/\sigma_T$, as might be expected to arise due to fermi motion and whose magnitude is within the error-bars of recent experiment, is shown to lead to a monotonic {\it decrease} in the ratio…
The QCD coupling alpha_s is determined from a combined analysis of experimental e+e- and e-p jet data confronted to theoretical predictions of the energy evolution of the parton-to-hadron fragmentation functions (FFs) moments…
In this paper, we show that Low Rank Adaptation (LoRA) as originally introduced in Hu et al. (2021) leads to suboptimal finetuning of models with large width (embedding dimension). This is due to the fact that adapter matrices A and B in…
It has been pointed out that the recent BaBar data on the pi gamma^* -> gamma transition form factor F_{pi gamma}(Q^2) at low (high) momentum transfer squared Q^2 indicate an asymptotic (flat) pion distribution amplitude. These seemingly…
Using an analytical parameterization for the behavior of the x slope of the structure function F_2 at small x in perturbative QCD, at the leading twist approximation of the Wilson operator product expansion, and applying a flat initial…