Related papers: A comment on the LLA method, the kT jet algorithm …
Existing methods for adapting large language models (LLMs) to new tasks are not suited to multi-task adaptation because they modify all the model weights -- causing destructive interference between tasks. The resulting effects, such as…
We discuss the high energy asymptotics in the next-to-leading (NLO) BFKL equation. We find a general solution for Green functions and consider two properties of the NLO BFKL kernel: running QCD coupling and large NLO corrections to the…
Model compression methods are used to reduce the computation and energy requirements for Large Language Models (LLMs). Quantization Aware Training (QAT), an effective model compression method, is proposed to reduce performance degradation…
We present the full next-to-leading order (NLO) prediction for the jet-gap-jet cross section at the LHC within the BFKL approach. We implement, for the first time, the NLO impact factors in the calculation of the cross section. We provide…
There is a substantial body of literature examining the mathematical reasoning capabilities of large language models (LLMs), particularly their performance on precise arithmetic operations in autoregressive architectures. However, their…
We compare the abilities of the cluster-type jet algorithm, KtJet, and a mid-point iterating cone algorithm to reconstruct the top mass at the LHC. We discuss the information contained in the merging scales of cluster-type algorithms, and…
Large language models (LLMs) equipped with chain-of-thoughts (CoT) prompting have shown significant multi-step reasoning capabilities in factual content like mathematics, commonsense, and logic. However, their performance in narrative…
Large language models (LLMs) have made significant advancements in natural language processing (NLP). Broad corpora capture diverse patterns but can introduce irrelevance, while focused corpora enhance reliability by reducing misleading…
Large Language Models (LLMs) have demonstrated great capabilities across diverse natural language tasks; yet their ability to solve abstraction and optimization problems with constraints remains scarcely explored. In this paper, we…
Large Language Models (LLMs) have shown remarkable progress across domains, yet their ability to perform inductive reasoning - inferring latent rules from sparse examples - remains limited. It is often assumed that chain-of-thought (CoT)…
A factorization formalism for jet processes involving massive colored particles such as the top quark is developed, extending earlier results for the massless case. The factorization of soft emissions from the underlying hard process is…
We show that next-leading logarithmic (NLL) Balitsky-Fadin-Kuraev-Lipatov (BFKL) effects can be tested by the forward-jet cross sections recently measured at HERA. For d\sigma/dx, the NLL corrections are small which confirms the stability…
The cognitive mechanism by which Large Language Models (LLMs) solve mathematical problems remains a widely debated and unresolved issue. Currently, there is little interpretable experimental evidence that connects LLMs' problem-solving with…
There has recently been much interest in analytical computations of jet mass distributions with and without vetos on additional jet activity [1-6]. An important issue affecting such calculations, particularly at next-to-leading logarithmic…
The state explosion problem and the exponentially computational complexity restrict the further applications of LTL model checking. To this end, this study tries to seek an acceptable approximate solution for LTL model checking by…
Chain-of-thought (CoT) reasoning has become the standard paradigm for enabling Large Language Models (LLMs) to solve complex problems. However, recent studies reveal a sharp performance drop in reasoning hop generalization scenarios, where…
We discuss the QCD corrections to the large-$m_t$ electroweak contributions to $\Delta r$ and to the process $Z\to b \bar b$ as two of the most representative examples. This needs the construction of an effetive field theory below the top…
The hallucination issue is recognized as a fundamental deficiency of large language models (LLMs), especially when applied to fields such as finance, education, and law. Despite the growing concerns, there has been a lack of empirical…
Robotic path planning problems are often NP-hard, and practical solutions typically rely on approximation algorithms with provable performance guarantees for general cases. While designing such algorithms is challenging, formally proving…
Linear and non-linear QCD evolutions at high energy suffer from severe issues related to convergence, due to higher order corrections enhanced by large double and single transverse logarithms. We resum double logarithms to all orders by…