Chelsea Finn
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Chelsea Finn has a total of 87 co-authors. Here are the top co-authors by number of shared publications:
- Sergey Levine
University of California, Berkeley
4 shared publications - Brijen Thananjeyan
Department of Electrical Engineering and Computer Sciences University of California Berkeley United States
3 shared publications Anikait Singh
UC Berkeley
2 shared publications- Archit Sharma
Stanford University
2 shared publications - Ashwin Balakrishna
Google DeepMind, Santa Clara, CA, USA
2 shared publications Aviral Kumar
UC Berkeley
2 shared publications
Latest Publications
- 12 September 2026
- Published by Springer Nature
- p. 37-55
Scaling Verification Can Be More Effective than Scaling Policy Learning for Vision-Language-Action Alignment
- 12 September 2026
- Published by Springer Nature
- p. 37-55
Scaling Verification Can Be More Effective than Scaling Policy Learning for Vision-Language-Action Alignment
- Open Access
Self-Guided Action Diffusion
Recent works have shown the promise of inference-time search over action samples for improving generative robot policies. In particular, optimizing cross-chunk coherence via bidirectional decoding has proven effective in boosting the consistency and reactivity of diffusion policies. However, this approach remains computationally expensive as the diversity of sampled actions grows. In this paper, we introduce self-guided action diffusion, a more efficient variant of bidirectional decoding tailored for diffusion-based policies. At the core of our method is to guide the proposal distribution at each diffusion step based on the prior decision. Experiments in simulation tasks show that the proposed self-guidance enables near-optimal performance at negligible inference cost. Notably, under a tight sampling budget, our method achieves up to 70% higher success rates than existing counterparts on challenging dynamic tasks. See project website at https://rhea-mal.github.io/selfgad.github.io.
Open AccessSelf-Guided Action Diffusion
Recent works have shown the promise of inference-time search over action samples for improving generative robot policies. In particular, optimizing cross-chunk coherence via bidirectional decoding has proven effective in boosting the consistency and reactivity of diffusion policies. However, this approach remains computationally expensive as the diversity of sampled actions grows. In this paper, we introduce self-guided action diffusion, a more efficient variant of bidirectional decoding tailored for diffusion-based policies. At the core of our method is to guide the proposal distribution at each diffusion step based on the prior decision. Experiments in simulation tasks show that the proposed self-guidance enables near-optimal performance at negligible inference cost. Notably, under a tight sampling budget, our method achieves up to 70% higher success rates than existing counterparts on challenging dynamic tasks. See project website at https://rhea-mal.github.io/selfgad.github.io.
- Open Access
- 11 October 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-19
Clarify: Improving Model Robustness With Natural Language Corrections
The standard way to teach models is by feeding them lots of data. However, this approach often teaches models incorrect ideas because they pick up on misleading signals in the data. To prevent such misconceptions, we must necessarily provide additional information beyond the training data. Prior methods incorporate additional instance-level supervision, such as labels for misleading features or additional labels for debiased data. However, such strategies require a large amount of labeler effort. We hypothesize that people are good at providing textual feedback at the concept level, a capability that existing teaching frameworks do not leverage. We propose Clarify, a novel interface and method for interactively correcting model misconceptions. Through Clarify, users need only provide a short text description of a model’s consistent failure patterns. Then, in an entirely automated way, we use such descriptions to improve the training process. Clarify is the first end-to-end system for user model correction. Our user studies show that non-expert users can successfully describe model misconceptions via Clarify, leading to increased worst-case performance in two datasets. We additionally conduct a case study on a large-scale image dataset, ImageNet, using Clarify to find and rectify 31 novel hard subpopulations.
- 11 October 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-19
Open AccessClarify: Improving Model Robustness With Natural Language Corrections
The standard way to teach models is by feeding them lots of data. However, this approach often teaches models incorrect ideas because they pick up on misleading signals in the data. To prevent such misconceptions, we must necessarily provide additional information beyond the training data. Prior methods incorporate additional instance-level supervision, such as labels for misleading features or additional labels for debiased data. However, such strategies require a large amount of labeler effort. We hypothesize that people are good at providing textual feedback at the concept level, a capability that existing teaching frameworks do not leverage. We propose Clarify, a novel interface and method for interactively correcting model misconceptions. Through Clarify, users need only provide a short text description of a model’s consistent failure patterns. Then, in an entirely automated way, we use such descriptions to improve the training process. Clarify is the first end-to-end system for user model correction. Our user studies show that non-expert users can successfully describe model misconceptions via Clarify, leading to increased worst-case performance in two datasets. We additionally conduct a case study on a large-scale image dataset, ImageNet, using Clarify to find and rectify 31 novel hard subpopulations.
- Open Access
- 7 March 2024
- Published by Association for Computing Machinery (ACM)
- p. 736-742
A Fast and Accurate Machine Learning Autograder for the Breakout Assignment
In this paper, we detail the successful deployment of a machine learning autograder that significantly decreases the grading labor required in the Breakout computer science assignment. This assignment - which tasks students with programming a game consisting of a controllable paddle and a ball that bounces off the paddle to break bricks - is popular for engaging students with introductory computer science concepts, but creates a large grading burden. Due to the game's interactive nature, grading defies traditional unit tests and instead typically requires 8+ minutes of manually playing each student's game to search for bugs. This amounts to 45+ hours of grading in a standard course offering and prevents further widespread adoption of the assignment. Our autograder alleviates this burden by playing each student's game with a reinforcement learning agent and providing videos of discovered bugs to instructors. In an A/B test with manual grading, we find that our human-in-the-loop AI autograder reduces grading time by 44%, while slightly improving grading accuracy by 6%, ultimately saving roughly 30 hours over our deployment in two offerings of the assignment. Our results further suggest the practicality of grading other interactive assignments (e.g., other games or building websites) via similar machine learning techniques. Live demo at https://ezliu.github.io/breakoutgrader.
- 7 March 2024
- Published by Association for Computing Machinery (ACM)
- p. 736-742
Open AccessA Fast and Accurate Machine Learning Autograder for the Breakout Assignment
In this paper, we detail the successful deployment of a machine learning autograder that significantly decreases the grading labor required in the Breakout computer science assignment. This assignment - which tasks students with programming a game consisting of a controllable paddle and a ball that bounces off the paddle to break bricks - is popular for engaging students with introductory computer science concepts, but creates a large grading burden. Due to the game's interactive nature, grading defies traditional unit tests and instead typically requires 8+ minutes of manually playing each student's game to search for bugs. This amounts to 45+ hours of grading in a standard course offering and prevents further widespread adoption of the assignment. Our autograder alleviates this burden by playing each student's game with a reinforcement learning agent and providing videos of discovered bugs to instructors. In an A/B test with manual grading, we find that our human-in-the-loop AI autograder reduces grading time by 44%, while slightly improving grading accuracy by 6%, ultimately saving roughly 30 hours over our deployment in two offerings of the assignment. Our results further suggest the practicality of grading other interactive assignments (e.g., other games or building websites) via similar machine learning techniques. Live demo at https://ezliu.github.io/breakoutgrader.
- Top Cited PapersOpen Access
- 1 January 2024
- Published by Association for Computational Linguistics (ACL)
- p. 4998-5017
Disentangling Length from Quality in Direct Preference Optimization
Ryan Park, Rafael Rafailov, Stefano Ermon, Chelsea Finn. Findings of the Association for Computational Linguistics ACL 2024. 2024.
- 1 January 2024
- Published by Association for Computational Linguistics (ACL)
- p. 4998-5017
Top Cited PapersOpen AccessDisentangling Length from Quality in Direct Preference Optimization
Ryan Park, Rafael Rafailov, Stefano Ermon, Chelsea Finn. Findings of the Association for Computational Linguistics ACL 2024. 2024.
- Open Access
Pre-Training for Robots: Offline RL Enables Learning New Tasks from a Handful of Trials
Progress in deep learning highlights the tremendous potential of utilizing diverse robotic datasets for attaining effective generalization and makes it enticing to consider leveraging broad datasets for attaining robust generalization in robotic learning as well. However, in practice, we often want to learn a new skill in a new environment that is unlikely to be contained in the prior data. Therefore we ask: how can we leverage existing diverse offline datasets in combination with small amounts of task-specific data to solve new tasks, while still enjoying the generalization benefits of training on large amounts of data? In this paper, we demonstrate that end-to-end offline RL can be an effective approach for doing this, without the need for any representation learning or vision-based pre-training. We present pre-training for robots (PTR), a framework based on offline RL that attempts to effectively learn new tasks by combining pre-training on existing robotic datasets with rapid fine-tuning on a new task, with as few as 10 demonstrations. PTR utilizes an existing offline RL method, conservative Q-learning (CQL), but extends it to include several crucial design decisions that enable PTR to actually work and outperform a variety of prior methods. To our knowledge, PTR is the first RL method that succeeds at learning new tasks in a new domain on a real WidowX robot with as few as 10 task demonstrations, by effectively leveraging an existing dataset of diverse multi-task robot data collected in a variety of toy kitchens. We also demonstrate that PTR can enable effective autonomous fine-tuning and improvement in a handful of trials, without needing any demonstrations. An accompanying overview video can be found in the supplementary material and at thi URL: https://sites.google.com/view/ptr-final/
Open AccessPre-Training for Robots: Offline RL Enables Learning New Tasks from a Handful of Trials
Progress in deep learning highlights the tremendous potential of utilizing diverse robotic datasets for attaining effective generalization and makes it enticing to consider leveraging broad datasets for attaining robust generalization in robotic learning as well. However, in practice, we often want to learn a new skill in a new environment that is unlikely to be contained in the prior data. Therefore we ask: how can we leverage existing diverse offline datasets in combination with small amounts of task-specific data to solve new tasks, while still enjoying the generalization benefits of training on large amounts of data? In this paper, we demonstrate that end-to-end offline RL can be an effective approach for doing this, without the need for any representation learning or vision-based pre-training. We present pre-training for robots (PTR), a framework based on offline RL that attempts to effectively learn new tasks by combining pre-training on existing robotic datasets with rapid fine-tuning on a new task, with as few as 10 demonstrations. PTR utilizes an existing offline RL method, conservative Q-learning (CQL), but extends it to include several crucial design decisions that enable PTR to actually work and outperform a variety of prior methods. To our knowledge, PTR is the first RL method that succeeds at learning new tasks in a new domain on a real WidowX robot with as few as 10 task demonstrations, by effectively leveraging an existing dataset of diverse multi-task robot data collected in a variety of toy kitchens. We also demonstrate that PTR can enable effective autonomous fine-tuning and improvement in a handful of trials, without needing any demonstrations. An accompanying overview video can be found in the supplementary material and at thi URL: https://sites.google.com/view/ptr-final/
- Open Access
- 8 August 2023
- Published by Association for Computing Machinery (ACM)
- p. 287-296
Self-Destructing Models: Increasing the Costs of Harmful Dual Uses of Foundation Models
A growing ecosystem of large, open-source foundation models has reduced the labeled data and technical expertise necessary to apply machine learning to many new problems. Yet foundation models pose a clear dual-use risk, indiscriminately reducing the costs of building both harmful and beneficial machine learning systems. Policy tools such as restricted model access and export controls are the primary methods currently used to mitigate such dual-use risks. In this work, we review potential safe-release strategies and argue that both policymakers and AI researchers would benefit from fundamentally new technologies enabling more precise control over the downstream usage of open-source foundation models. We propose one such approach: the task blocking paradigm, in which foundation models are trained with an additional mechanism to impede adaptation to harmful tasks without sacrificing performance on desirable tasks. We call the resulting models self-destructing models, inspired by mechanisms that prevent adversaries from using tools for harmful purposes. We present an algorithm for training self-destructing models leveraging techniques from meta-learning and adversarial learning, which we call meta-learned adversarial censoring (MLAC). In a small-scale experiment, we show MLAC can largely prevent a BERT-style model from being re-purposed to perform gender identification without harming the model’s ability to perform profession classification.
- 8 August 2023
- Published by Association for Computing Machinery (ACM)
- p. 287-296
Open AccessSelf-Destructing Models: Increasing the Costs of Harmful Dual Uses of Foundation Models
A growing ecosystem of large, open-source foundation models has reduced the labeled data and technical expertise necessary to apply machine learning to many new problems. Yet foundation models pose a clear dual-use risk, indiscriminately reducing the costs of building both harmful and beneficial machine learning systems. Policy tools such as restricted model access and export controls are the primary methods currently used to mitigate such dual-use risks. In this work, we review potential safe-release strategies and argue that both policymakers and AI researchers would benefit from fundamentally new technologies enabling more precise control over the downstream usage of open-source foundation models. We propose one such approach: the task blocking paradigm, in which foundation models are trained with an additional mechanism to impede adaptation to harmful tasks without sacrificing performance on desirable tasks. We call the resulting models self-destructing models, inspired by mechanisms that prevent adversaries from using tools for harmful purposes. We present an algorithm for training self-destructing models leveraging techniques from meta-learning and adversarial learning, which we call meta-learned adversarial censoring (MLAC). In a small-scale experiment, we show MLAC can largely prevent a BERT-style model from being re-purposed to perform gender identification without harming the model’s ability to perform profession classification.
- Open Access
Waypoint-Based Imitation Learning for Robotic Manipulation
While imitation learning methods have seen a resurgent interest for robotic manipulation, the well-known problem of compounding errors continues to afflict behavioral cloning (BC). Waypoints can help address this problem by reducing the horizon of the learning problem for BC, and thus, the errors compounded over time. However, waypoint labeling is underspecified, and requires additional human supervision. Can we generate waypoints automatically without any additional human supervision? Our key insight is that if a trajectory segment can be approximated by linear motion, the endpoints can be used as waypoints. We propose Automatic Waypoint Extraction (AWE) for imitation learning, a preprocessing module to decompose a demonstration into a minimal set of waypoints which when interpolated linearly can approximate the trajectory up to a specified error threshold. AWE can be combined with any BC algorithm, and we find that AWE can increase the success rate of state-of-the-art algorithms by up to 25% in simulation and by 4-28% on real-world bimanual manipulation tasks, reducing the decision making horizon by up to a factor of 10. Videos and code are available at https://lucys0.github.io/awe/
Open AccessWaypoint-Based Imitation Learning for Robotic Manipulation
While imitation learning methods have seen a resurgent interest for robotic manipulation, the well-known problem of compounding errors continues to afflict behavioral cloning (BC). Waypoints can help address this problem by reducing the horizon of the learning problem for BC, and thus, the errors compounded over time. However, waypoint labeling is underspecified, and requires additional human supervision. Can we generate waypoints automatically without any additional human supervision? Our key insight is that if a trajectory segment can be approximated by linear motion, the endpoints can be used as waypoints. We propose Automatic Waypoint Extraction (AWE) for imitation learning, a preprocessing module to decompose a demonstration into a minimal set of waypoints which when interpolated linearly can approximate the trajectory up to a specified error threshold. AWE can be combined with any BC algorithm, and we find that AWE can increase the success rate of state-of-the-art algorithms by up to 25% in simulation and by 4-28% on real-world bimanual manipulation tasks, reducing the decision making horizon by up to a factor of 10. Videos and code are available at https://lucys0.github.io/awe/
- Open Access
- 10 July 2023
- Published by Robotics: Science and Systems Foundation
Pre-Training for Robots: Offline RL Enables Learning New Tasks in a Handful of Trials
- 10 July 2023
- Published by Robotics: Science and Systems Foundation
Open AccessPre-Training for Robots: Offline RL Enables Learning New Tasks in a Handful of Trials
- Open Access
Self-Improving Robots: End-to-End Autonomous Visuomotor Reinforcement Learning
In imitation and reinforcement learning, the cost of human supervision limits the amount of data that robots can be trained on. An aspirational goal is to construct self-improving robots: robots that can learn and improve on their own, from autonomous interaction with minimal human supervision or oversight. Such robots could collect and train on much larger datasets, and thus learn more robust and performant policies. While reinforcement learning offers a framework for such autonomous learning via trial-and-error, practical realizations end up requiring extensive human supervision for reward function design and repeated resetting of the environment between episodes of interactions. In this work, we propose MEDAL++, a novel design for self-improving robotic systems: given a small set of expert demonstrations at the start, the robot autonomously practices the task by learning to both do and undo the task, simultaneously inferring the reward function from the demonstrations. The policy and reward function are learned end-to-end from high-dimensional visual inputs, bypassing the need for explicit state estimation or task-specific pre-training for visual encoders used in prior work. We first evaluate our proposed algorithm on a simulated non-episodic benchmark EARL, finding that MEDAL++ is both more data efficient and gets up to 30% better final performance compared to state-of-the-art vision-based methods. Our real-robot experiments show that MEDAL++ can be applied to manipulation problems in larger environments than those considered in prior work, and autonomous self-improvement can improve the success rate by 30-70% over behavior cloning on just the expert data. Code, training and evaluation videos along with a brief overview is available at: https://architsharma97.github.io/self-improving-robots/
Open AccessSelf-Improving Robots: End-to-End Autonomous Visuomotor Reinforcement Learning
In imitation and reinforcement learning, the cost of human supervision limits the amount of data that robots can be trained on. An aspirational goal is to construct self-improving robots: robots that can learn and improve on their own, from autonomous interaction with minimal human supervision or oversight. Such robots could collect and train on much larger datasets, and thus learn more robust and performant policies. While reinforcement learning offers a framework for such autonomous learning via trial-and-error, practical realizations end up requiring extensive human supervision for reward function design and repeated resetting of the environment between episodes of interactions. In this work, we propose MEDAL++, a novel design for self-improving robotic systems: given a small set of expert demonstrations at the start, the robot autonomously practices the task by learning to both do and undo the task, simultaneously inferring the reward function from the demonstrations. The policy and reward function are learned end-to-end from high-dimensional visual inputs, bypassing the need for explicit state estimation or task-specific pre-training for visual encoders used in prior work. We first evaluate our proposed algorithm on a simulated non-episodic benchmark EARL, finding that MEDAL++ is both more data efficient and gets up to 30% better final performance compared to state-of-the-art vision-based methods. Our real-robot experiments show that MEDAL++ can be applied to manipulation problems in larger environments than those considered in prior work, and autonomous self-improvement can improve the success rate by 30-70% over behavior cloning on just the expert data. Code, training and evaluation videos along with a brief overview is available at: https://architsharma97.github.io/self-improving-robots/