Related papers: Testing the efficiency of different improvement pr…

This paper presents a mathematical approach for improving the performance of a control system by modifying the time delay at certain operating conditions. This approach converts a continuous time loop into a discrete time loop. The formula…

Large Language Models (LLMs) have demonstrated remarkable capabilities in software engineering, yet comprehensive benchmarks covering diverse SE activities remain limited. We present a multi-task evaluation of 11 state-of-the-art LLMs…

We study the empirical scaling laws of a family of encoder-decoder autoregressive transformer models on the task of joint motion forecasting and planning in the autonomous driving domain. Using a 500 thousand hours driving dataset, we…

We extend our earlier work on the massive $O(N)$ nonlinear sigma model to other observables. We derive expressions at leading order in the large $N$ expansion at all orders in the loop expansion for the decay constant, vacuum expectation…

This paper contributes to the literature on treatment effects estimation with machine learning inspired methods by studying the performance of different estimators based on the Lasso. Building on recent work in the field of high-dimensional…

Self-improvement is a significant techniques within the realm of large language model (LLM), aiming to enhance the LLM performance without relying on external data. Despite its significance, generally how LLM performances evolve during the…

We apply the finite size scaling analysis to the derivative of the density of the effective action for the lattice U(1) pure gauge theory in an external constant magnetic field. We found the presence of a continuous phase transition.…

We classify and re-examine some of the current approaches to improve the performance-computes trade-off of language models, including (1) non-causal models (such as masked language models), (2) extension of batch length with efficient…

Data-centric learning seeks to improve model performance from the perspective of data quality, and has been drawing increasing attention in the machine learning community. Among its key tools, influence functions provide a powerful…

We study one-loop effective action of hypermultiplet theory coupled to external N=2 vector multiplet. We formulate this theory in N=1 superspace and develop a general approach to constructing derivative expansion of the effective action…

This thesis establishes new quantitative records in several problems of incidence geometry and growth. After the necessary background in Chapters 1, 2 and 3, the following results are proven. Chapter 4 gives new results in the incidence…

In this work, we conduct an assessment of the optimization capabilities of LLMs across various tasks and data sizes. Each of these tasks corresponds to unique optimization domains, and LLMs are required to execute these tasks with…

Majority voting over multiple LLM attempts improves mathematical reasoning, but correlated errors limit the effective sample size. A natural fix is to assign different reasoning strategies to different voters. The approach, Diverse Prompt…

Large Language Model (LLM)-based Multi-Agent Systems (MAS) have emerged as a powerful paradigm for tackling complex, multi-step tasks across diverse domains. However, despite their impressive capabilities, MAS remain susceptible to…

Underlying relationships among Multi-Agent Systems (MAS) in hazardous scenarios can be represented as Game-theoretic models. This paper proposes a new hierarchical network-based model called Game-theoretic Utility Tree (GUT), which…

Our aim is to present a new model which encompasses pace optimization and motor control effort for a runner on a fixed distance. We see that for long races, the long term behaviour is well approximated by a turnpike problem. We provide…

In order to support the formal renormalization group arguments that the fixed point action of an asymptotically free model gives cut--off independent physical predictions in 1--loop perturbation theory, we calculate the finite volume…

Large Transformer-based language models are pre-trained on corpora of varying sizes, for a different number of steps and with different batch sizes. At the same time, more fundamental components, such as the pre-training objective or…

In this paper, we determine which non-random sampling of fixed size gives the best linear predictor of the sum of a finite spatial population. We employ different multiscale superpopulation models and use the minimum mean-squared error as…

Modelling growth in student achievement is a significant challenge in the field of education. Understanding how interventions or experiences such as part-time work can influence this growth is also important. Traditional methods like…