A5049073875- Teaching and Learning Programming
- Software Engineering Techniques and Practices
- Online Learning and Analytics
- Advanced Software Engineering Methodologies
- Software Engineering Research
- E-Learning and Knowledge Management
- Service-Oriented Architecture and Web Services
- Reinforcement Learning in Robotics
- Topic Modeling
- Logic, programming, and type systems
- Adversarial Robustness in Machine Learning
- Artificial Intelligence in Law
- Business Process Modeling and Analysis
- Formal Methods in Verification
- Model-Driven Software Engineering Techniques
- Natural Language Processing Techniques
- Distributed and Parallel Computing Systems
- Scientific Computing and Data Management
- Computability, Logic, AI Algorithms
- Information Systems Education and Curriculum Development
- Ethics and Social Impacts of AI
- Law, AI, and Intellectual Property
- Experimental Learning in Engineering
- Evolutionary Algorithms and Applications
- Parallel Computing and Optimization Techniques
Princeton University
1974-2025
Center for Information Technology
2024-2025
Princeton Public Schools
2024-2025
Stanford University
2020-2024
King's College School
2024
Mount Sinai Medical Center
2023
The King's College
2022
McGill University
2015-2019
Intelligent Machines (Sweden)
2017
Princeton Plasma Physics Laboratory
2009-2016
No abstract available.
AI is undergoing a paradigm shift with the rise of models (e.g., BERT, DALL-E, GPT-3) that are trained on broad data at scale and adaptable to wide range downstream tasks. We call these foundation underscore their critically central yet incomplete character. This report provides thorough account opportunities risks models, ranging from capabilities language, vision, robotics, reasoning, human interaction) technical principles(e.g., model architectures, training procedures, data, systems,...
In recent years, significant progress has been made in solving challenging problems across various domains using deep reinforcement learning (RL). Reproducing existing work and accurately judging the improvements offered by novel methods is vital to sustaining this progress. Unfortunately, reproducing results for state-of-the-art RL seldom straightforward. particular, non-determinism standard benchmark environments, combined with variance intrinsic methods, can make reported tough interpret....
This degree work aims to explore the use of synthetic financial time series generated by a Generative Adversarial Neural Networks (GAN) model train Deep Reinforcement Learning algorithm that executes buy and sell actions for stock in Standard & Poor's 500 index.For implementation study, we used CRISP methodology proposed IBM, understanding first business theory necessary develop models, continue with exploration knowledge available data matched objectives project.In this paper, procedure...
Large language models (LLMs) have been shown to be able perform new tasks based on a few demonstrations or natural instructions. While these capabilities led widespread adoption, most LLMs are developed by resource-rich organizations and frequently kept from the public. As step towards democratizing this powerful technology, we present BLOOM, 176B-parameter open-access model designed built thanks collaboration of hundreds researchers. BLOOM is decoder-only Transformer that was trained ROOTS...
A different way to execute pure LISP programs is presented. It delays the evaluation of parameters and list structures without ever having perform more steps than usual method. Although central idea can be found in earlier work this paper interest since it treats a rather well-known language works out an algorithm which avoids full substitution. partial correctness proof using Scott-Strachey semantics sketched later section.
Accurate reporting of energy and carbon usage is essential for understanding the potential climate impacts machine learning research. We introduce a framework that makes this easier by providing simple interface tracking realtime consumption emissions, as well generating standardized online appendices. Utilizing framework, we create leaderboard efficient reinforcement algorithms to incentivize responsible research in area an example other areas learning. Finally, based on case studies using...
Policy gradient methods in reinforcement learning have become increasingly prevalent for state-of-the-art performance continuous control tasks. Novel typically benchmark against a few key algorithms such as deep deterministic policy gradients and trust region optimization. As such, it is important to present use consistent baselines experiments. However, this can be difficult due general variance the algorithms, hyper-parameter tuning, environment stochasticity. We investigate discuss:...
During the past decade, several areas of speech and language understanding have witnessed substantial breakthroughs from use data-driven models. In area dialogue systems, trend is less obvious, most practical systems are still built through significant engineering expert knowledge. Nevertheless, recent results suggest that approaches feasible quite promising. To facilitate research in this area, we carried out a wide survey publicly available datasets suitable for learning systems. We...
With the recent wave of progress in artificial intelligence (AI) has come a growing awareness large-scale impacts AI systems, and recognition that existing regulations norms industry academia are insufficient to ensure responsible development. In order for developers earn trust from system users, customers, civil society, governments, other stakeholders they building responsibly, will need make verifiable claims which can be held accountable. Those outside given organization also effective...
The advent of large language models (LLMs) and their adoption by the legal community has given rise to question: what types reasoning can LLMs perform? To enable greater study this question, we present LegalBench: a collaboratively constructed benchmark consisting 162 tasks covering six different reasoning. LegalBench was built through an interdisciplinary process, in which collected designed hand-crafted professionals. Because these subject matter experts took leading role construction,...
Existing foundation models are trained on copyrighted material. Deploying these can pose both legal and ethical risks when data creators fail to receive appropriate attribution or compensation. In the United States several other countries, content may be used build without incurring liability due fair use doctrine. However, there is a caveat: If model produces output that similar data, particularly in scenarios affect market of no longer apply model. this work, we emphasize not guaranteed,...
Warning: this paper contains data, prompts, and model outputs that are offensive in nature. Recently, there has been a surge of interest integrating vision into Large Language Models (LLMs), exemplified by Visual (VLMs) such as Flamingo GPT-4. This sheds light on the security safety implications trend. First, we underscore continuous high-dimensional nature visual input makes it weak link against adversarial attacks, representing an expanded attack surface vision-integrated LLMs. Second,...
The use of dialogue systems as a medium for human-machine interaction is an increasingly prevalent paradigm. A growing number conversation strategies that are learned from large datasets. There well documented instances where interactions with these system have resulted in biased or even offensive conversations due to the data-driven training process. Here, we highlight potential ethical issues arise research, including: implicit biases systems, rise adversarial examples, sources privacy...
During the past decade, several areas of speech and language understanding have witnessed substantial breakthroughs from use data-driven models. In area dialogue systems, trend is less obvious, most practical systems are still built through significant engineering expert knowledge. Nevertheless, recent results suggest that approaches feasible quite promising. To facilitate research in this area, we carried out a wide survey publicly available datasets suitable for learning systems. We...
Despite its importance to experimental design, statistical power (the probability that, given a real effect, an experiment will reject the null hypothesis) has largely been ignored by NLP community. Underpowered experiments make it more difficult discern difference between noise and meaningful model improvements, increase chances of exaggerated findings. By meta-analyzing set existing papers datasets, we characterize typical for variety settings conclude that underpowered are common in...
While self-supervised learning has made rapid advances in natural language processing, it remains unclear when researchers should engage resource-intensive domain-specific pretraining (domain pretraining). The law, puzzlingly, yielded few documented instances of substantial gains to domain spite the fact that legal is widely seen be unique. We hypothesize these existing results stem from NLP tasks are too easy and fail meet conditions for can help. To address this, we first present CaseHOLD...
Language models (LMs) are becoming the foundation for almost all major language technologies, but their capabilities, limitations, and risks not well understood. We present Holistic Evaluation of Models (HELM) to improve transparency models. First, we taxonomize vast space potential scenarios (i.e. use cases) metrics desiderata) that interest LMs. Then select a broad subset based on coverage feasibility, noting what's missing or underrepresented (e.g. question answering neglected English...
Is AI set to redefine the legal profession? We argue that this claim is not supported by current evidence. dive into AI's increasingly prevalent roles in three types of tasks: information processing; tasks involving creativity, reasoning, or judgment; and predictions about future. find ease evaluating applications varies greatly across tasks, based on identifying correct answers observability relevant task at hand. Tasks would lead most significant changes profession are also ones prone...
Many of the process synchronization problems studied in literature are form a conjunction finitely many conditions type “process $p_i$ blocks $p_j$”. Such may be expressed as directed graphs whose nodes represent processes and where there is an edge from node i to j if only $p_j$. We characterize class which correspond system synchronizing primitives Vantilborgh van Lamsweerde terms normal representation present efficient algorithm for determining whether arbitrary graph this class.