Osbert Bastani

papers

993

total citations

papers (20)

Program Synthesis Guided Reinforcement Learning for Partially Observed Environments

NEURIPS 2021arXiv

citations

Zeroth-Order Fine-Tuning of LLMs with Transferable Static Sparsity

ICLR 2025

citations

Alignment of Large Language Models with Constrained Learning

NEURIPS 2025arXiv

citations

Stochastic Online Conformal Prediction with Semi-Bandit Feedback

ICML 2025arXiv

citations

Stochastic Bandits with ReLU Neural Networks

ICML 2024arXiv

citations

Offline Goal-Conditioned Reinforcement Learning via $f$-Advantage Regression

NEURIPS 2022

citations

Neurosymbolic Deep Generative Models for Sequence Data with Relational Constraints

NEURIPS 2022

citations

Angelic Patches for Improving Third-Party Object Detector Performance

CVPR 2023

citations

Osbert Bastani

papers (20)

Eureka: Human-Level Reward Design via Coding Large Language Models

Conservative Offline Distributional Reinforcement Learning

Compositional Reinforcement Learning from Logical Specifications

Practical Adversarial Multivalid Conformal Prediction

Vision Language Models are In-Context Value Learners

Likelihood-Based Diverse Sampling for Trajectory Forecasting

Evaluating the Diversity and Quality of LLM Generated Content

Neurosymbolic Transformers for Multi-Agent Communication

Regret Bounds for Risk-Sensitive Reinforcement Learning

PAC Prediction Sets for Meta-Learning

Learning Models for Actionable Recourse

PAC Prediction Sets Under Label Shift

Program Synthesis Guided Reinforcement Learning for Partially Observed Environments

Zeroth-Order Fine-Tuning of LLMs with Transferable Static Sparsity

Alignment of Large Language Models with Constrained Learning

Stochastic Online Conformal Prediction with Semi-Bandit Feedback

Stochastic Bandits with ReLU Neural Networks

Offline Goal-Conditioned Reinforcement Learning via $f$-Advantage Regression

Neurosymbolic Deep Generative Models for Sequence Data with Relational Constraints

Angelic Patches for Improving Third-Party Object Detector Performance

papers (20)

Eureka: Human-Level Reward Design via Coding Large Language Models

Conservative Offline Distributional Reinforcement Learning

Compositional Reinforcement Learning from Logical Specifications

Practical Adversarial Multivalid Conformal Prediction

Vision Language Models are In-Context Value Learners

Likelihood-Based Diverse Sampling for Trajectory Forecasting

Evaluating the Diversity and Quality of LLM Generated Content

Neurosymbolic Transformers for Multi-Agent Communication

Regret Bounds for Risk-Sensitive Reinforcement Learning

PAC Prediction Sets for Meta-Learning

Learning Models for Actionable Recourse

PAC Prediction Sets Under Label Shift

Program Synthesis Guided Reinforcement Learning for Partially Observed Environments

Zeroth-Order Fine-Tuning of LLMs with Transferable Static Sparsity

Alignment of Large Language Models with Constrained Learning

Stochastic Online Conformal Prediction with Semi-Bandit Feedback

Stochastic Bandits with ReLU Neural Networks

Offline Goal-Conditioned Reinforcement Learning via $f$-Advantage Regression

Neurosymbolic Deep Generative Models for Sequence Data with Relational Constraints

Angelic Patches for Improving Third-Party Object Detector Performance