Laura Smith

ORCID: 0000-0002-7890-497X
Publications
Citations
Views
---
Saved
---
About
Contact & Profiles
Research Areas
  • Reinforcement Learning in Robotics
  • X-ray Diffraction in Crystallography
  • Crystallization and Solubility Studies
  • Vehicle emissions and performance
  • Advanced Combustion Engine Technologies
  • Crystallography and molecular interactions
  • Molecular Junctions and Nanostructures
  • Employment and Welfare Studies
  • Participatory Visual Research Methods
  • Community Health and Development
  • Robotic Locomotion and Control
  • Advanced Chemical Physics Studies
  • Sex work and related issues
  • Metal and Thin Film Mechanics
  • Education Systems and Policy
  • Urban, Neighborhood, and Segregation Studies
  • Electron and X-Ray Spectroscopy Techniques
  • Biodiesel Production and Applications
  • Traffic and Road Safety
  • Microstructure and mechanical properties
  • CCD and CMOS Imaging Sensors
  • Mobile Crowdsensing and Crowdsourcing
  • CO2 Reduction Techniques and Catalysts
  • Educational and Psychological Assessments
  • Mining and Resource Management

University of Illinois Urbana-Champaign
1960-2024

University of Illinois System
1959-2024

University of Georgia
2024

University of California, Berkeley
2023

BAE Systems (United States)
2022

Queen's University
2022

Gemeente Maastricht
2022

Columbia University
2010

Barnard College
2003

Kaiser Family Foundation
2001

Participatory action research (PAR) is a methodological stance that researchers can find both inspiring and daunting. Community-based PAR offers platform by which social scientists contribute to the democratization of knowledge its production, but also requires they go beyond conventional roles procedures interact with community co-researchers in ways may leave university-based feeling exposed rudderless. In this article, authors present episodes from three different projects illustrate some...

10.1177/1476750310366043 article EN Action Research 2010-06-30

Preregistration is the act of formally documenting a research plan before collecting (or at least analysing) data. It allows those reading final report to know which aspects study were decided sight data, and added later. This enables informed evaluation severity with scientific claims have been tested. We, as British Neuropsychological Society Open Science Group, conducted survey explore awareness adoption open science practices within our field. Neuropsychology involves relatively rare or...

10.31234/osf.io/p267j preprint EN 2025-01-04

Agile skills learned with our proposed method enable the A1 robot to jump repeatedly (left) and walk a goal location

10.15607/rss.2023.xix.051 article EN 2023-07-10

Deep reinforcement learning is a promising approach to policies in uncontrolled environments that do not require domain knowledge. Unfortunately, due sample inefficiency, deep RL applications have primarily focused on simulated environments. In this work, we demonstrate the recent advancements machine algorithms and libraries combined with carefully tuned robot controller lead quadruped locomotion only 20 minutes real world. We evaluate our several indoor outdoor terrains which are known be...

10.48550/arxiv.2208.07860 preprint EN other-oa arXiv (Cornell University) 2022-01-01

Photoassisted catalysis using Ni complexes is an emerging field for cross-coupling reactions in organic synthesis. However, the mechanism by which light enables and enhances reactivity of these often remains elusive. Although optical techniques have been widely used to study ground excited states photocatalysts, they lack specificity interrogate electronic structural changes at specific atoms. Herein, we report metal-specific studies transient L- K-edge X-ray absorption spectroscopy a...

10.1021/acs.jpclett.4c00226 article EN cc-by The Journal of Physical Chemistry Letters 2024-05-01

Conveying complex objectives to reinforcement learning (RL) agents can often be difficult, involving meticulous design of reward functions that are sufficiently informative yet easy enough provide. Human-in-the-loop RL methods allow practitioners instead interactively teach through tailored feedback; however, such approaches have been challenging scale since human feedback is very expensive. In this work, we aim make process more sample- and feedback-efficient. We present an off-policy,...

10.48550/arxiv.2106.05091 preprint EN other-oa arXiv (Cornell University) 2021-01-01

Ferrocene is one of the most common electron donors, and mapping its ligand-field excited states critical to designing donor-acceptor (D-A) molecules with long-lived charge transfer states. Although

10.1021/jacs.4c05646 article EN cc-by Journal of the American Chemical Society 2024-07-25

Youth participatory action research (YPAR) projects offer young people the opportunity to increase their sociocultural awareness, critical thinking abilities, and sense of agency within a collaborative group experience. Thus far, however, such have been primarily province educators social psychologists, not substantively explored as basis for school counseling interventions. This article suggests initiation exploration framework existing ecological justice models practice, presents an...

10.5330/prsc.14.2.m62r11337332gt54 article EN Professional School Counseling 2010-12-01

Preregistration is the act of formally documenting a research plan before collecting (or at least analysing) data. It allows those reading final report to know which aspects study were decided sight data, and added later. This enables informed evaluation severity with scientific claims have been tested. We, as British Neuropsychological Society Open Science Group, conducted survey explore awareness adoption open science practices within our field. Neuropsychology involves relatively rare or...

10.31234/osf.io/p267j_v1 preprint EN 2025-01-04

Preregistration is the act of formally documenting a research plan before collecting (or at least analysing) data. It allows those reading final report to know which aspects study were decided sight data, and added later. This enables informed evaluation severity with scientific claims have been tested. We, as British Neuropsychological Society Open Science Group, conducted survey explore awareness adoption open science practices within our field. Neuropsychology involves relatively rare or...

10.31234/osf.io/p267j_v2 preprint EN 2025-03-18

Sample efficiency and exploration remain major challenges in online reinforcement learning (RL). A powerful approach that can be applied to address these issues is the inclusion of offline data, such as prior trajectories from a human expert or sub-optimal policy. Previous methods have relied on extensive modifications additional complexity ensure effective use this data. Instead, we ask: simply apply existing off-policy leverage data when online? In work, demonstrate answer yes; however,...

10.48550/arxiv.2302.02948 preprint EN cc-by arXiv (Cornell University) 2023-01-01

Counselors are often challenged to address issues of social justice in the counseling context, and they must be deliberate innovative their attempts respond. will required relate considerations practices theoretical foundations these practices; then operationalize an approach that suits particular practice setting. The authors present early results attempt meet this challenge.

10.1002/j.2161-1882.2003.tb00222.x article EN Journal of College Counseling 2003-03-01

Reinforcement learning (RL) requires access to a reward function that incentivizes the right behavior, but these are notoriously hard specify for complex tasks. Preference-based RL provides an alternative: policies using teacher's preferences without pre-defined rewards, thus overcoming concerns associated with engineering. However, it is difficult quantify progress in preference-based due lack of commonly adopted benchmark. In this paper, we introduce B-Pref: benchmark specially designed...

10.48550/arxiv.2111.03026 preprint EN other-oa arXiv (Cornell University) 2021-01-01

Meta-reinforcement learning (RL) methods can meta-train policies that adapt to new tasks with orders of magnitude less data than standard RL, but meta-training itself is costly and time-consuming. If we on offline data, then reuse the same static dataset, labeled once rewards for different tasks, a variety at meta-test time. Although this capability would make meta-RL practical tool real-world use, presents additional challenges beyond online or RL settings. Meta-RL learns an exploration...

10.48550/arxiv.2107.03974 preprint EN other-oa arXiv (Cornell University) 2021-01-01

Ferrocene is one of the most common electron donors, and mapping its ligand-field excited states critical to designing donor-acceptor (D-A) molecules with long-lived charge transfer states. Although 3(d-d) are commonly invoked in photophysics ferrocene complexes, mention high-spin 5(d-d) state scarce. Here, we provide clear evidence formation a bimetallic D-A molecule, ferrocenyl cobaltocenium hexafluorophosphate ([FcCc]PF6). Femtosecond optical transient absorption (OTA) spectroscopy...

10.26434/chemrxiv-2024-8mxvd preprint EN cc-by-nc-nd 2024-06-03

Replicating human-like dexterity in robot hands represents one of the largest open problems robotics. Reinforcement learning is a promising approach that has achieved impressive progress last few years; however, class it typically addressed corresponds to rather narrow definition as compared human capabilities. To address this gap, we investigate piano-playing, skill challenges even limits dexterity, means test high-dimensional control, and which requires high spatial temporal precision,...

10.48550/arxiv.2304.04150 preprint EN cc-by arXiv (Cornell University) 2023-01-01

10.1177/002205741708601120 article EN Journal of Education 1917-09-01

Skill discovery methods enable agents to learn diverse emergent behaviors without explicit rewards. To make learned skills useful for unknown downstream tasks, obtaining a semantically repertoire of is essential. While some approaches introduce discriminator distinguish and others aim increase state coverage, no existing work directly addresses the "semantic diversity" skills. We hypothesize that leveraging semantic knowledge large language models (LLMs) can lead us improve diversity...

10.48550/arxiv.2406.06615 preprint EN arXiv (Cornell University) 2024-06-07

Ferrocene is one of the most common electron donors, and mapping its ligand-field excited states critical to designing donor-acceptor (D-A) molecules with long-lived charge transfer states. Although 3(d-d) are commonly invoked in photophysics ferrocene complexes, mention high-spin 5(d-d) state scarce. Here, we provide clear evidence formation a bimetallic D-A molecule, ferrocenyl cobaltocenium hexafluorophosphate ([FcCc]PF6). Femtosecond optical transient absorption (OTA) spectroscopy...

10.26434/chemrxiv-2024-8mxvd-v2 preprint EN 2024-07-09

Abstract Excessive clutter can create a barrier for older adults to maintain functional lives in their homes; however, interventions hoarding are lengthy, expensive, and often unavailable, particularly rural-dwelling adults. Using the Rightsizing Program as an example, this presentation will highlight synergy between Extension academic faculty promote aging place across diverse states. The is collaboration specialists at Mississippi State University (MSU) of Georgia (UGA) psychology MSU with...

10.1093/geroni/igae098.1700 article EN cc-by Innovation in Aging 2024-12-01

Cities across Canada are aging—fast. As the “baby boom” generation enters older adulthood, cities facing a dramatic demographic shift. However, it is unclear whether in prepared to meet needs of growing adult population. Through case study one Canada’s demographically oldest cities, Victoria, BC, we assess age-friendliness local plans and policies. Specifically, conduct plan evaluation content analysis city’s Official Community Plan supporting statutory determine city truly planning for an...

10.24908/cppapc.v2022i1.15291 article EN cc-by Canadian Planning and Policy / Aménagement et politique au Canada 2022-06-30
Coming Soon ...