PyTorchKR Papers

최신 논문

총 658편
태그·기간 등 상세 필터

카테고리 탐색

로그인 후 이 태그 구독

arXiv:2501.18177

Investigating Tax Evasion Emergence Using Dual Large Language Model and Deep Reinforcement Learning Powered Agent-based Simulation

Tax evasion, usually the largest component of an informal economy, is a persistent challenge over history with significant socio-economic implications. Many socio-economic studies investigate its dynamics, including influencing factors, the role and influence of taxation policies, and the prediction of the tax evasion volume over time. These studies assumed such behavior is given, as observed in the real world, neglecting the "big bang" of such activity in a population. To this end, computational economy studies adopted developments in computer simulations, in general, and recent innovations in artificial intelligence (AI), in particular, to simulate and study informal economy appearance in various socio-economic settings. This study presents a novel computational framework to examine the dynamics of tax evasion and the emergence of informal economic activity. Employing an agent-based simulation powered by Large Language Models and Deep Reinforcement Learning, the framework is uniquely designed to allow informal economic behaviors to emerge organically, without presupposing their existence or explicitly signaling agents about the possibility of evasion. This provides a rigorous approach for exploring the socio-economic determinants of compliance behavior. The experimental design, comprising model validation and exploratory phases, demonstrates the framework's robustness in replicating theoretical economic behaviors. Findings indicate that individual personality traits, external narratives, enforcement probabilities, and the perceived efficiency of public goods provision significantly influence both the timing and extent of informal economic activity. The results underscore that efficient public goods provision and robust enforcement mechanisms are complementary; neither alone is sufficient to curtail informal activity effectively.

arXiv:2501.15458

Amortized Safe Active Learning for Real-Time Data Acquisition: Pretrained Neural Policies from Simulated Nonparametric Functions

Safe active learning (AL) is a sequential scheme for learning unknown systems while respecting safety constraints during data acquisition. Existing methods often rely on Gaussian processes (GPs) to model the task and safety constraints, requiring repeated GP updates and constrained acquisition optimization-incurring in significant computations which are challenging for real-time decision-making. We propose an amortized safe AL framework that replaces expensive online computations with a pretrained neural policy. Inspired by recent advances in amortized Bayesian experimental design, we turn GPs into a pretraining simulator. We train our policy prior to the AL deployment on simulated nonparametric functions, using Fourier feature-based GP sampling and a differentiable, safety-aware acquisition objective. At deployment, our policy selects safe and informative queries via a single forward pass, eliminating the need for GP inference or constrained optimization. This leads to substantial speed improvements while preserving safety and learning quality. Our framework is modular and can be adapted to unconstrained, time-sensitive AL tasks by omitting the safety requirement.

arXiv:2501.10385공식 코드 · Papers with Code

Autonomous Microscopy Experiments through Large Language Model Agents

The emergence of large language models (LLMs) has accelerated the development of self-driving laboratories (SDLs) for materials research. Despite their transformative potential, current SDL implementations rely on rigid, predefined protocols that limit their adaptability to dynamic experimental scenarios across different labs. A significant challenge persists in measuring how effectively AI agents can replicate the adaptive decision-making and experimental intuition of expert scientists. Here, we introduce AILA (Artificially Intelligent Lab Assistant), a framework that automates atomic force microscopy (AFM) through LLM-driven agents. Using AFM as an experimental testbed, we develop AFMBench-a comprehensive evaluation suite that challenges AI agents based on language models like GPT-4o and GPT-3.5 to perform tasks spanning the scientific workflow: from experimental design to results analysis. Our systematic assessment shows that state-of-the-art language models struggle even with basic tasks such as documentation retrieval, leading to a significant decline in performance in multi-agent coordination scenarios. Further, we observe that LLMs exhibit a tendency to not adhere to instructions or even divagate to additional tasks beyond the original request, raising serious concerns regarding safety alignment aspects of AI agents for SDLs. Finally, we demonstrate the application of AILA on increasingly complex experiments open-ended experiments: automated AFM calibration, high-resolution feature detection, and mechanical property measurement. Our findings emphasize the necessity for stringent benchmarking protocols before deploying AI agents as laboratory assistants across scientific disciplines.

arXiv:2501.10030공식 코드 · Papers with Code

Informativity Conditions for Multiple Signals: Properties, Experimental Design, and Applications

Recent studies highlight the importance of persistently exciting condition in single signal sequence for model identification and data-driven control methodologies. However, maintaining prolonged excitation in control signals introduces significant challenges, as continuous excitation can reduce the lifetime of mechanical devices. In this paper, we introduce three informativity conditions for various types of multi-signal data, each augmented by weight factors. We explore the interrelations between these conditions and their rank properties in linear time-invariant systems. Furthermore, we introduce open-loop experimental design methods tailored to each of the three conditions, which can synthesize the required excitation conditions either offline or online, even in the presence of limited information within each signal segment. We demonstrate the effectiveness of these informativity conditions in least-squares identification. Additionally, all three conditions can extend Willems' fundamental lemma and are utilized to assess the properties of the system. Illustrative examples confirm that these conditions yield satisfactory outcomes in both least-squares identification and the construction of data-driven controllers.

arXiv:2501.06112

Optimizing Experiments for Accurate Battery Circuit Parameters Estimation: Reduction and Adjustment of Frequency Set Used in Electrochemical Impedance Spectroscopy

In this paper, we study a suitable experimental design of electrochemical impedance spectroscopy (EIS) to reduce the number of frequency points while not significantly affecting the uncertainties of the estimated cell's equivalent circuit model (ECM) parameters. It is based on an E-optimal experimental design that aims to maximize the information about the ECM parameters collected by EIS measurements and, at the same time, minimize the overall uncertainty. In a numerical experiment, we first analyze to which extent reducing the number of measurement points at low frequencies affects the uncertainty of the estimated parameters. Secondly, we show that applying the frequency adjustments can lead to the same or even improved global uncertainty of ECM parameter estimates as with a higher number of measurements. This is numerically verified through a case study using the ECM parameters of a commercial battery cell.

arXiv:2501.06506

Resource Allocation under the Latin Square Constraint

A Latin square is an $n \times n$ matrix filled with $n$ distinct symbols, each of which appears exactly once in each row and exactly once in each column. We introduce a problem of allocating $n$ indivisible items among $n$ agents over $n$ rounds while satisfying the Latin square constraint. This constraint ensures that each agent receives no more than one item per round and receives each item at most once. Each agent has an additive valuation on the item--round pairs. Real-world applications like scheduling, resource management, and experimental design require the Latin square constraint to satisfy fairness or balancedness in allocation. Our goal is to find a partial or complete allocation that maximizes the sum of the agents' valuations (utilitarian social welfare) or the minimum of the agents' valuations (egalitarian social welfare). For the problem of maximizing utilitarian social welfare, we prove NP-hardness even when the valuations are binary additive. We then provide $(1-1/e)$ and $(1-1/e)/4$-approximation algorithms for partial and complete settings, respectively. Additionally, we present fixed-parameter tractable (FPT) algorithms with respect to the order of Latin square and the optimum value for both partial and complete settings. For the problem of maximizing egalitarian social welfare, we establish that deciding whether the optimum value is at most $1$ or at least $2$ is NP-hard for both the partial and complete settings, even when the valuations are binary. Furthermore, we demonstrate that checking the existence of a complete allocation that satisfies each of envy-free, proportional, equitable, envy-free up to any good, proportional up to any good, or equitable up to any good is NP-hard, even when the valuations are identical.

arXiv:2501.02673

Exploring a Datasets Statistical Effect Size Impact on Model Performance, and Data Sample-Size Sufficiency

Having a sufficient quantity of quality data is a critical enabler of training effective machine learning models. Being able to effectively determine the adequacy of a dataset prior to training and evaluating a model's performance would be an essential tool for anyone engaged in experimental design or data collection. However, despite the need for it, the ability to prospectively assess data sufficiency remains an elusive capability. We report here on two experiments undertaken in an attempt to better ascertain whether or not basic descriptive statistical measures can be indicative of how effective a dataset will be at training a resulting model. Leveraging the effect size of our features, this work first explores whether or not a correlation exists between effect size, and resulting model performance (theorizing that the magnitude of the distinction between classes could correlate to a classifier's resulting success). We then explore whether or not the magnitude of the effect size will impact the rate of convergence of our learning rate, (theorizing again that a greater effect size may indicate that the model will converge more rapidly, and with a smaller sample size needed). Our results appear to indicate that this is not an effective heuristic for determining adequate sample size or projecting model performance, and therefore that additional work is still needed to better prospectively assess adequacy of data.

arXiv:2501.01540공식 코드 · Papers with Code

BoxingGym: Benchmarking Progress in Automated Experimental Design and Model Discovery

Understanding the world and explaining it with scientific theories is a central aspiration of artificial intelligence research. Proposing theories, designing experiments to test them, and then revising them based on data are fundamental to scientific discovery. Despite the significant promise of LLM-based scientific agents, no benchmarks systematically test LLM's ability to propose scientific models, collect experimental data, and revise them in light of new data. We introduce BoxingGym, a benchmark with 10 environments for systematically evaluating both experimental design (e.g. collecting data to test a scientific theory) and model discovery (e.g. proposing and revising scientific theories). To enable tractable and quantitative evaluation, we implement each environment as a generative probabilistic model with which a scientific agent can run interactive experiments. These probabilistic models are drawn from various real-world scientific domains ranging from psychology to ecology. To quantitatively evaluate a scientific agent's ability to collect informative experimental data, we compute the expected information gain (EIG), an information-theoretic quantity which measures how much an experiment reduces uncertainty about the parameters of a generative model. A good scientific theory is a concise and predictive explanation. Therefore, to quantitatively evaluate model discovery, we ask a scientific agent to explain their model and then assess whether this explanation enables another scientific agent to make reliable predictions about this environment. In addition to this explanation-based evaluation, we compute standard model evaluation metrics such as prediction errors. We find that current LLMs, such as GPT-4o, struggle with both experimental design and model discovery. We find that augmenting the LLM-based agent with an explicit statistical model does not reliably improve these results.

arXiv:2501.01283

A Systematic Computational Framework for Practical Identifiability Analysis in Mathematical Models Arising from Biology

Practical identifiability is a critical concern in data-driven modeling of mathematical systems. In this paper, we propose a novel framework for practical identifiability analysis to evaluate parameter identifiability in mathematical models of biological systems. Starting with a rigorous mathematical definition of practical identifiability, we demonstrate its equivalence to the invertibility of the Fisher Information Matrix. Our framework establishes the relationship between practical identifiability and coordinate identifiability, introducing a novel metric that simplifies and accelerates the evaluation of parameter identifiability compared to the profile likelihood method. Additionally, we introduce new regularization terms to address non-identifiable parameters, enabling uncertainty quantification and improving model reliability. To guide experimental design, we present an optimal data collection algorithm that ensures all model parameters are practically identifiable. Applications to Hill functions, neural networks, and dynamic biological models demonstrate the feasibility and efficiency of the proposed computational framework in uncovering critical biological processes and identifying key observable variables.

arXiv:2501.00972

Optimal Sampling for Generalized Linear Model under Measurement Constraint with Surrogate Variables

Measurement-constrained datasets, often encountered in semi-supervised learning, arise when data labeling is costly, time-intensive, or hindered by confidentiality or ethical concerns, resulting in a scarcity of labeled data. In certain cases, surrogate variables are accessible across the entire dataset and can serve as approximations to the true response variable; however, these surrogates often contain measurement errors and thus cannot be directly used for accurate prediction. We propose an optimal sampling strategy that effectively harnesses the available information from surrogate variables. This approach provides consistent estimators under the assumption of a generalized linear model, achieving theoretically lower asymptotic variance than existing optimal sampling algorithms that do not use surrogate data information. By employing the A-optimality criterion from optimal experimental design, our strategy maximizes statistical efficiency. Numerical studies demonstrate that our approach surpasses existing optimal sampling methods, exhibiting reduced empirical mean squared error and enhanced robustness in algorithmic performance. These findings highlight the practical advantages of our strategy in scenarios where measurement constraints exist and surrogates are available.

arXiv:2412.20429

Multi-Scenario Reasoning: Unlocking Cognitive Autonomy in Humanoid Robots for Multimodal Understanding

To improve the cognitive autonomy of humanoid robots, this research proposes a multi-scenario reasoning architecture to solve the technical shortcomings of multi-modal understanding in this field. It draws on simulation based experimental design that adopts multi-modal synthesis (visual, auditory, tactile) and builds a simulator "Maha" to perform the experiment. The findings demonstrate the feasibility of this architecture in multimodal data. It provides reference experience for the exploration of cross-modal interaction strategies for humanoid robots in dynamic environments. In addition, multi-scenario reasoning simulates the high-level reasoning mechanism of the human brain to humanoid robots at the cognitive level. This new concept promotes cross-scenario practical task transfer and semantic-driven action planning. It heralds the future development of self-learning and autonomous behavior of humanoid robots in changing scenarios.

arXiv:2412.18329

Comprehensive Analysis and Experimental Design of High-Gain DC-DC Boost Converter Topologies

Global demand for clean and eco-friendly energy sources has inspired decades of far-reaching research in power generation from renewable energy sources. Solar cells, wind, and tidal sources are limited in output power generation compared to the fast-rising power requirements of most industrial applications. Besides, the efficiency of conventional DC-DC boost converters is significantly low due to the presence of parasitic elements culminating in switching losses. This study presents three high-gain boost converter topologies for optimizing the limited voltage generation by solar Photovoltaic cells. The three converters are realized based on a modification of a classical Cuk converter. Simply put, the first proposed converter is realized by the inclusion of one capacitor and one inductor to the classical topology. Similarly, the addition of 3 capacitors, 2 diodes, and one inductor leads to the practical realization of the second proposed topology. Similarly, the third proposed topology consists of additional 3 diodes and 3 capacitors. Based on this estimation, the first, second, and third proposed high gain modified Cuk converter topologies generate output voltages 10 times, 20 times, and 29 times the input voltage respectively when the switching device is gated at 90% duty ratio. Theoretical/mathematical analysis validates the Lt-spice numerical simulation results of all the proposed converters. Furthermore, experimental prototype results were compared with Lt-spice estimation to determine the accuracy of the converters.

arXiv:2412.12936

A simple DNN regression for the chemical composition in essential oil

Although experimental design and methodological surveys for mono-molecular activity/property has been extensively investigated, those for chemical composition have received little attention, with the exception of a few prior studies. In this study, we configured three simple DNN regressors to predict essential oil property based on chemical composition. Despite showing overfitting due to the small size of dataset, all models were trained effectively in this study.

arXiv:2412.11104공식 코드 · Papers with Code

ABC3: Active Bayesian Causal Inference with Cohn Criteria in Randomized Experiments

In causal inference, randomized experiment is a de facto method to overcome various theoretical issues in observational study. However, the experimental design requires expensive costs, so an efficient experimental design is necessary. We propose ABC3, a Bayesian active learning policy for causal inference. We show a policy minimizing an estimation error on conditional average treatment effect is equivalent to minimizing an integrated posterior variance, similar to Cohn criteria \citep{cohn1994active}. We theoretically prove ABC3 also minimizes an imbalance between the treatment and control groups and the type 1 error probability. Imbalance-minimizing characteristic is especially notable as several works have emphasized the importance of achieving balance. Through extensive experiments on real-world data sets, ABC3 achieves the highest efficiency, while empirically showing the theoretical results hold.

arXiv:2412.11526

Probabilities-Informed Machine Learning

Machine learning (ML) has emerged as a powerful tool for tackling complex regression and classification tasks, yet its success often hinges on the quality of training data. This study introduces an ML paradigm inspired by domain knowledge of the structure of output function, akin to physics-informed ML, but rooted in probabilistic principles rather than physical laws. The proposed approach integrates the probabilistic structure of the target variable (such as its cumulative distribution function) into the training process. This probabilistic information is obtained from historical data or estimated using structural reliability methods during experimental design. By embedding domain-specific probabilistic insights into the learning process, the technique enhances model accuracy and mitigates risks of overfitting and underfitting. Applications in regression, image denoising, and classification demonstrate the approach's effectiveness in addressing real-world problems.

arXiv:2412.05516

Measuring Consumer Sensitivity to Audio Advertising: A Long-Run Field Experiment on Pandora Internet Radio

A randomized experiment with almost 35 million Pandora listeners enables us to measure the sensitivity of consumers to advertising, an important topic of study in the era of ad-supported digital content provision. The experiment randomized listeners into nine treatment groups, each of which received a different level of audio advertising interrupting their music listening, with the highest treatment group receiving more than twice as many ads as the lowest treatment group. By maintaining consistent treatment assignment for 21 months, we measure long-run demand effects and find ad-load sensitivity three times greater than what we would have obtained from a month-long experiment. We show the negative impact on the number of hours listened, days listened, and probability of listening at all in the final month. Using an experimental design that separately varies the number of commercial interruptions per hour and the number of ads per commercial interruption, we find that listeners primarily respond to the total number of ads per hour, with a slight preference for more frequent but shorter ad breaks. Lastly, we find that increased ad load led to an increase in the number of paid ad-free subscriptions to Pandora. Importantly, we show that observational methods often lead to biased or even directionally incorrect estimates of these effects, highlighting the value of experimental data.

arXiv:2412.09429

From Intention To Implementation: Automating Biomedical Research via LLMs

Conventional biomedical research is increasingly labor-intensive due to the exponential growth of scientific literature and datasets. Artificial intelligence (AI), particularly Large Language Models (LLMs), has the potential to revolutionize this process by automating various steps. Still, significant challenges remain, including the need for multidisciplinary expertise, logicality of experimental design, and performance measurements. This paper introduces BioResearcher, the first end-to-end automated system designed to streamline the entire biomedical research process involving dry lab experiments. BioResearcher employs a modular multi-agent architecture, integrating specialized agents for search, literature processing, experimental design, and programming. By decomposing complex tasks into logically related sub-tasks and utilizing a hierarchical learning approach, BioResearcher effectively addresses the challenges of multidisciplinary requirements and logical complexity. Furthermore, BioResearcher incorporates an LLM-based reviewer for in-process quality control and introduces novel evaluation metrics to assess the quality and automation of experimental protocols. BioResearcher successfully achieves an average execution success rate of 63.07% across eight previously unmet research objectives. The generated protocols, on average, outperform typical agent systems by 22.0% on five quality metrics. The system demonstrates significant potential to reduce researchers' workloads and accelerate biomedical discoveries, paving the way for future innovations in automated research systems.

arXiv:2412.08098

What You See Is Not Always What You Get: An Empirical Study of Code Comprehension by Large Language Models

Recent studies have demonstrated outstanding capabilities of large language models (LLMs) in software engineering tasks, including code generation and comprehension. While LLMs have shown significant potential in assisting with coding, it is perceived that LLMs are vulnerable to adversarial attacks. In this paper, we investigate the vulnerability of LLMs to imperceptible attacks, where hidden character manipulation in source code misleads LLMs' behaviour while remaining undetectable to human reviewers. We devise these attacks into four distinct categories and analyse their impacts on code analysis and comprehension tasks. These four types of imperceptible coding character attacks include coding reordering, invisible coding characters, code deletions, and code homoglyphs. To comprehensively benchmark the robustness of current LLMs solutions against the attacks, we present a systematic experimental evaluation on multiple state-of-the-art LLMs. Our experimental design introduces two key performance metrics, namely model confidence using log probabilities of response, and the response correctness. A set of controlled experiments are conducted using a large-scale perturbed and unperturbed code snippets as the primary prompt input. Our findings confirm the susceptibility of LLMs to imperceptible coding character attacks, while different LLMs present different negative correlations between perturbation magnitude and performance. These results highlight the urgent need for robust LLMs capable of manoeuvring behaviours under imperceptible adversarial conditions. We anticipate this work provides valuable insights for enhancing the security and trustworthiness of LLMs in software engineering applications.

arXiv:2412.08085공식 코드 · Papers with Code

Non-Myopic Multi-Objective Bayesian Optimization

We consider the problem of finite-horizon sequential experimental design to solve multi-objective optimization (MOO) of expensive black-box objective functions. This problem arises in many real-world applications, including materials design, where we have a small resource budget to make and evaluate candidate materials in the lab. We solve this problem using the framework of Bayesian optimization (BO) and propose the first set of non-myopic methods for MOO problems. Prior work on non-myopic BO for single-objective problems relies on the Bellman optimality principle to handle the lookahead reasoning process. However, this principle does not hold for most MOO problems because the reward function needs to satisfy some conditions: scalar variable, monotonicity, and additivity. We address this challenge by using hypervolume improvement (HVI) as our scalarization approach, which allows us to use a lower-bound on the Bellman equation to approximate the finite-horizon using a batch expected hypervolume improvement (EHVI) acquisition function (AF) for MOO. Our formulation naturally allows us to use other improvement-based scalarizations and compare their efficacy to HVI. We derive three non-myopic AFs for MOBO: 1) the Nested AF, which is based on the exact computation of the lower bound, 2) the Joint AF, which is a lower bound on the nested AF, and 3) the BINOM AF, which is a fast and approximate variant based on batch multi-objective acquisition functions. Our experiments on multiple diverse real-world MO problems demonstrate that our non-myopic AFs substantially improve performance over the existing myopic AFs for MOBO.

arXiv:2412.07444

MO-IOHinspector: Anytime Benchmarking of Multi-Objective Algorithms using IOHprofiler

Benchmarking is one of the key ways in which we can gain insight into the strengths and weaknesses of optimization algorithms. In sampling-based optimization, considering the anytime behavior of an algorithm can provide valuable insights for further developments. In the context of multi-objective optimization, this anytime perspective is not as widely adopted as in the single-objective context. In this paper, we propose a new software tool which uses principles from unbounded archiving as a logging structure. This leads to a clearer separation between experimental design and subsequent analysis decisions. We integrate this approach as a new Python module into the IOHprofiler framework and demonstrate the benefits of this approach by showcasing the ability to change indicators, aggregations, and ranking procedures during the analysis pipeline.