Alert button
Picture for Michael Carbin

Michael Carbin

Alert button

The Cost of Down-Scaling Language Models: Fact Recall Deteriorates before In-Context Learning

Oct 07, 2023
Tian Jin, Nolan Clement, Xin Dong, Vaishnavh Nagarajan, Michael Carbin, Jonathan Ragan-Kelley, Gintare Karolina Dziugaite

Figure 1 for The Cost of Down-Scaling Language Models: Fact Recall Deteriorates before In-Context Learning
Figure 2 for The Cost of Down-Scaling Language Models: Fact Recall Deteriorates before In-Context Learning
Figure 3 for The Cost of Down-Scaling Language Models: Fact Recall Deteriorates before In-Context Learning
Figure 4 for The Cost of Down-Scaling Language Models: Fact Recall Deteriorates before In-Context Learning
Viaarxiv icon

Turaco: Complexity-Guided Data Sampling for Training Neural Surrogates of Programs

Sep 21, 2023
Alex Renda, Yi Ding, Michael Carbin

Figure 1 for Turaco: Complexity-Guided Data Sampling for Training Neural Surrogates of Programs
Figure 2 for Turaco: Complexity-Guided Data Sampling for Training Neural Surrogates of Programs
Figure 3 for Turaco: Complexity-Guided Data Sampling for Training Neural Surrogates of Programs
Figure 4 for Turaco: Complexity-Guided Data Sampling for Training Neural Surrogates of Programs
Viaarxiv icon

Computably Continuous Reinforcement-Learning Objectives are PAC-learnable

Mar 19, 2023
Cambridge Yang, Michael Littman, Michael Carbin

Figure 1 for Computably Continuous Reinforcement-Learning Objectives are PAC-learnable
Viaarxiv icon

Acela: Predictable Datacenter-level Maintenance Job Scheduling

Dec 10, 2022
Yi Ding, Aijia Gao, Thibaud Ryden, Kaushik Mitra, Sukumar Kalmanje, Yanai Golany, Michael Carbin, Henry Hoffmann

Figure 1 for Acela: Predictable Datacenter-level Maintenance Job Scheduling
Figure 2 for Acela: Predictable Datacenter-level Maintenance Job Scheduling
Figure 3 for Acela: Predictable Datacenter-level Maintenance Job Scheduling
Figure 4 for Acela: Predictable Datacenter-level Maintenance Job Scheduling
Viaarxiv icon

Pruning's Effect on Generalization Through the Lens of Training and Regularization

Oct 25, 2022
Tian Jin, Michael Carbin, Daniel M. Roy, Jonathan Frankle, Gintare Karolina Dziugaite

Figure 1 for Pruning's Effect on Generalization Through the Lens of Training and Regularization
Figure 2 for Pruning's Effect on Generalization Through the Lens of Training and Regularization
Figure 3 for Pruning's Effect on Generalization Through the Lens of Training and Regularization
Figure 4 for Pruning's Effect on Generalization Through the Lens of Training and Regularization
Viaarxiv icon

SCOPE: Safe Exploration for Dynamic Computer Systems Optimization

Apr 22, 2022
Hyunji Kim, Ahsan Pervaiz, Henry Hoffmann, Michael Carbin, Yi Ding

Figure 1 for SCOPE: Safe Exploration for Dynamic Computer Systems Optimization
Figure 2 for SCOPE: Safe Exploration for Dynamic Computer Systems Optimization
Figure 3 for SCOPE: Safe Exploration for Dynamic Computer Systems Optimization
Figure 4 for SCOPE: Safe Exploration for Dynamic Computer Systems Optimization
Viaarxiv icon

Cello: Efficient Computer Systems Optimization with Predictive Early Termination and Censored Regression

Apr 11, 2022
Yi Ding, Alex Renda, Ahsan Pervaiz, Michael Carbin, Henry Hoffmann

Figure 1 for Cello: Efficient Computer Systems Optimization with Predictive Early Termination and Censored Regression
Figure 2 for Cello: Efficient Computer Systems Optimization with Predictive Early Termination and Censored Regression
Figure 3 for Cello: Efficient Computer Systems Optimization with Predictive Early Termination and Censored Regression
Figure 4 for Cello: Efficient Computer Systems Optimization with Predictive Early Termination and Censored Regression
Viaarxiv icon

Programming with Neural Surrogates of Programs

Dec 12, 2021
Alex Renda, Yi Ding, Michael Carbin

Figure 1 for Programming with Neural Surrogates of Programs
Figure 2 for Programming with Neural Surrogates of Programs
Figure 3 for Programming with Neural Surrogates of Programs
Figure 4 for Programming with Neural Surrogates of Programs
Viaarxiv icon

Reinforcement Learning for General LTL Objectives Is Intractable

Nov 24, 2021
Cambridge Yang, Michael Littman, Michael Carbin

Figure 1 for Reinforcement Learning for General LTL Objectives Is Intractable
Figure 2 for Reinforcement Learning for General LTL Objectives Is Intractable
Figure 3 for Reinforcement Learning for General LTL Objectives Is Intractable
Figure 4 for Reinforcement Learning for General LTL Objectives Is Intractable
Viaarxiv icon