Yuqing Du

yuqing_du (at) berkeley (dot) edu

Hi! I'm a Staff Research Scientist at Google DeepMind. Currently, I work on post-training and multimodal generation (Nano Banana, Imagen, Veo). I am broadly interested in human-AI interaction research, especially from a multimodal lens.

I received my PhD from UC Berkeley, advised by Professor Pieter Abbeel at the BAIR lab. My thesis, "Human-Centric Reward Design", focuses on using RL to enable AI agents to learn from and alongside people. During this time, I also dabbled in investigating if RL can improve reasoning in LLMs at FAIR and connecting generative models and robotics at DeepMind and Everyday Robots.

Before this, I received a B.A.Sc in Engineering Physics with a minor in Honours Mathematics from UBC. There I was advised by Professors Machiel Van der Loos and Elizabeth Croft at the CARIS Lab, working on human-robot interaction research.

papers

Learning to Model the World with Language

Jessy Lin, Yuqing Du, Olivia Watkins, Danijar Hafner, Pieter Abbeel, Dan Klein, and Anca Dragan

International Conference on Machine Learning (ICML) 2024
(Oral)

Abstract arXiv Website Code

To interact with humans in the world, agents need to understand the diverse types of language that people use, relate them to the visual world, and act based on them. While current agents learn to execute simple language instructions from task rewards, we aim to build agents that leverage diverse language that conveys general knowledge, describes the state of the world, provides interactive feedback, and more. Our key idea is that language helps agents predict the future: what will be observed, how the world will behave, and which situations will be rewarded. This perspective unifies language understanding with future prediction as a powerful self-supervised learning objective. We present Dynalang, an agent that learns a multimodal world model that predicts future text and image representations and learns to act from imagined model rollouts. Unlike traditional agents that use language only to predict actions, Dynalang acquires rich language understanding by using past language also to predict future language, video, and rewards. In addition to learning from online interaction in an environment, Dynalang can be pretrained on datasets of text, video, or both without actions or rewards. From using language hints in grid worlds to navigating photorealistic scans of homes, Dynalang utilizes diverse types of language to improve task performance, including environment descriptions, game rules, and instructions.
Teaching Large Language Models to Reason with Reinforcement Learning

Alexander Havrilla, Yuqing Du, Sharath Chandra Raparthy, Christoforos Nalmpantis, Jane Dwivedi-Yu, Maksym Zhuravinskyi, Eric Hambro, Sainbayar Sukhbaatar, and Roberta Raileanu

AI4MATH Workshop @ ICML 2024

Abstract arXiv

Reinforcement Learning from Human Feedback (RLHF) has emerged as a dominant approach for align- ing LLM outputs with human preferences. Inspired by the success of RLHF, we study the performance of multiple algorithms that learn from feedback (Expert Iteration, Proximal Policy Optimization (PPO), Return-Conditioned RL) on improving LLM reasoning capabilities. We investigate both sparse and dense rewards provided to the LLM both heuristically and via a learned reward model. We additionally start from multiple model sizes and initializations both with and without supervised fine-tuning (SFT) data. Overall, we find all algorithms perform comparably, with Expert Iteration performing best in most cases. Surprisingly, we find the sample complexity of Expert Iteration is similar to that of PPO, requiring at most on the order of 106 samples to converge from a pretrained checkpoint. We investigate why this is the case, concluding that during RL training models fail to explore significantly beyond solutions already produced by SFT models. Additionally, we discuss a trade off between maj@1 and pass@96 metric performance during SFT training and how conversely RL training improves both simultaneously. We then conclude by discussing the implications of our findings for RLHF and the future role of RL in LLM fine-tuning.
What can AI Learn from Human Exploration? Intrinsically-Motivated Humans and Agents in Open-World Exploration

Yuqing Du*, Eliza Kosoy*, Alyssa L. Dayan*, Maria Rufova, Pieter Abbeel, and Alison Gopnik

Intrinsically Motivated Open-Ended Learning Workshop @ NeurIPS 2023
(Spotlight)

Abstract PDF

What drives exploration? Understanding intrinsic motivation is a long-standing question in both cognitive science and artificial intelligence (AI); numerous ex- ploration objectives have been proposed and tested in human experiments and used to train reinforcement learning (RL) agents. However, experiments in the former are often in simplistic environments that do not capture the complexity of real world exploration. On the other hand, experiments in the latter use more complex environments, yet the trained RL agents fail to come close to human exploration efficiency. To study this gap, we propose a framework for directly com- paring human and agent exploration in an open-ended environment, Crafter [23]. We study how well commonly-proposed information theoretic intrinsic objectives relate to actual human and agent behaviours, finding that human and intrinsically- motivated RL agent exploration success consistently show positive correlation with Entropy and Empowerment. However, only human exploration shows significant correlation with Information Gain. In a preliminary analysis of verbalizations, we find that children’s verbalizations of goals show a strong positive correlation with Empowerment, suggesting that goal-setting may be an important aspect of efficient exploration.
A Study on Improving Reasoning in Language Models

Yuqing Du, Alexander Havrilla, Sainbayar Sukhbaatar, Pieter Abbeel, and Roberta Raileanu

I Can’t Believe It’s Not Better! (ICBINB) Workshop @ NeurIPS 2023

Abstract PDF

Accurately carrying out complex reasoning is a crucial component of deployable and reliable language models. While current language models can exhibit this capability with few-shot guidance, accurate reasoning is primarily restricted to larger model sizes. In this work, we explore methods for improving the reasoning capabilities of smaller language models which are more deployable than their larger counterparts. Specifically, we look at variations of supervised learning, online reinforcement learning with PPO, and distillation from larger models. Surprisingly, for reasoning tasks such as CommonsenseQA and GSM8K we find that simple filtered supervised learning often outperforms reward-conditioned super- vised learning, and that simple iterative supervised learning performs on par with online reinforcement learning.
DPOK: Reinforcement Learning for Fine-tuning Text-to-Image Diffusion Models

Ying Fan, Olivia Watkins, Yuqing Du, Hao Liu, Moonkyung Ryu, Craig Boutilier, Pieter Abbeel, Mohammad Ghavamzadeh, Kangwook Lee, and Kimin Lee

Neural Information Processing Systems (NeurIPS) 2023

Abstract arXiv Website

Learning from human feedback has been shown to improve text-to-image models. These techniques first learn a reward function that captures what humans care about in the task and then improve the models based on the learned reward function. Even though relatively simple approaches (e.g., rejection sampling based on reward scores) have been investigated, fine-tuning text-to-image models with the reward function remains challenging. In this work, we propose using online reinforcement learning (RL) to fine-tune text-to-image models. We focus on diffusion models, defining the fine-tuning task as an RL problem, and updating the pre-trained text-to-image diffusion models using policy gradient to maximize the feedback-trained reward. Our approach, coined DPOK, integrates policy optimization with KL regularization. We conduct an analysis of KL regularization for both RL fine-tuning and supervised fine-tuning. In our experiments, we show that DPOK is generally superior to supervised fine-tuning with respect to both image-text alignment and image quality.
Vision-Language Models as Success Detectors

Yuqing Du, Ksenia Konyushkova, Misha Denil, Akhil Raju, Jessica Landon, Felix Hill, Nando de Freitas, and Serkan Cabi

Conference on Lifelong Learning Agents (CoLLAs) 2023
(Oral)

Abstract arXiv

Detecting successful behaviour is crucial for training intelligent agents. As such, generalisable reward models are a prerequisite for agents that can learn to generalise their behaviour. In this work we focus on developing robust success detectors that leverage large, pretrained vision-language models (Flamingo, Alayrac et al. (2022)) and human reward annotations. Concretely, we treat success detection as a visual question answering (VQA) problem, denoted SuccessVQA. We study success detection across three vastly different domains: (i) interactive language-conditioned agents in a simulated household, (ii) real world robotic manipulation, and (iii) "in-the-wild" human egocentric videos. We investigate the generalisation properties of a Flamingo-based success detection model across unseen language and visual changes in the first two domains, and find that the proposed method is able to outperform bespoke reward models in out-of-distribution test scenarios with either variation. In the last domain of "in-the-wild" human videos, we show that success detection on unseen real videos presents an even more challenging generalisation task warranting future work. We hope our initial results encourage further work in real world success detection and reward modelling.
Aligning Text-to-Image Models using Human Feedback

Kimin Lee, Hao Liu, Moonkyung Ryu, Olivia Watkins, Yuqing Du, Craig Boutilier, Pieter Abbeel, Mohammad Ghavamzadeh, and Shixiang Shane Gu

Abstract arXiv

Deep generative models have shown impressive results in text-to-image synthesis. However, current text-to-image models often generate images that are inadequately aligned with text prompts. We propose a fine-tuning method for aligning such models using human feedback, comprising three stages. First, we collect human feedback assessing model output alignment from a set of diverse text prompts. We then use the human-labeled image-text dataset to train a reward function that predicts human feedback. Lastly, the text-to-image model is fine-tuned by maximizing reward-weighted likelihood to improve image-text alignment. Our method generates objects with specified colors, counts and backgrounds more accurately than the pre-trained model. We also analyze several design choices and find that careful investigations on such design choices are important in balancing the alignment-fidelity tradeoffs. Our results demonstrate the potential for learning from human feedback to significantly improve text-to-image models.
Guiding Pretraining in Reinforcement Learning with Large Language Models

Yuqing Du*, Olivia Watkins*, Zihan Wang, Cédric Colas, Trevor Darrell, Pieter Abbeel, Abhishek Gupta, and Jacob Andreas

International Conference on Machine Learning (ICML) 2023

Abstract arXiv Code

Reinforcement learning algorithms typically struggle in the absence of a dense, well-shaped reward function. Intrinsically motivated exploration methods address this limitation by rewarding agents for visiting novel states or transitions, but these methods offer limited benefits in large environments where most discovered novelty is irrelevant for downstream tasks. We describe a method that uses background knowledge from text corpora to shape exploration. This method, called ELLM (Exploring with LLMs) rewards an agent for achieving goals suggested by a language model prompted with a description of the agent's current state. By leveraging large-scale language model pretraining, ELLM guides agents toward human-meaningful and plausibly useful behaviors without requiring a human in the loop. We evaluate ELLM in the Crafter game environment and the Housekeep robotic simulator, showing that ELLM-trained agents have better coverage of common-sense behaviors during pretraining and usually match or improve performance on a range of downstream tasks.
Practical Visual Deep Imitation Learning via Task-Level Domain Consistency

Mohi Khansari, Daniel Ho, Yuqing Du, Armando Fuentes, Matthew Bennice, Nicolas Sievers, Sean Kirmani, Yunfei Bai, and Eric Jang

International Conference on Robotics and Automation (ICRA) 2023

Abstract arXiv

Recent work in visual end-to-end learning for robotics has shown the promise of imitation learning across a variety of tasks. Such approaches are however expensive both because they require large amounts of real world data and rely on time-consuming real-world evaluations to identify the best model for deployment. These challenges can be mitigated by using simulation evaluations to identify high performing policies. However, this introduces the well-known “reality gap” problem, where simulator inaccuracies decorrelate performance in simulation from that of reality. In this paper, we build on top of prior work in GAN-based domain adaptation and introduce the notion of a Task Consistency Loss (TCL), a self-supervised loss that encourages sim and real alignment both at the feature and action-prediction levels. We demonstrate the effectiveness of our approach by teaching a 9-DoF mobile manipulator to perform the challenging task of latched door opening purely from visual inputs such as RGB and depth images. We achieve 69% success across twenty seen and unseen meeting rooms using only ~ 16.2 hours of teleoperated demonstrations in sim and real. To the best of our knowledge, this is the first work to tackle latched door opening from a purely end-to-end learning approach, where the task of navigation and manipulation are jointly modeled by a single neural network.
Bayesian Imitation Learning for End-to-End Mobile Manipulation

Yuqing Du, Daniel Ho, Alexander A. Alemi, Eric Jang, and Mohi Khansari

International Conference on Machine Learning (ICML) 2022

Abstract arXiv

In this work we investigate and demonstrate benefits of a Bayesian approach to imitation learning from multiple sensor inputs, as applied to the task of opening office doors with a mobile manipulator. Augmenting policies with additional sensor inputs, such as RGB + depth cameras, is a straightforward approach to improving robot perception capabilities, especially for tasks that may favor different sensors in different situations. As we scale multi-sensor robotic learning to unstructured real-world settings (e.g. offices, homes) and more complex robot behaviors, we also increase reliance on simulators for cost, efficiency, and safety. Consequently, the sim-to-real gap across multiple sensor modalities also increases, making simulated validation more difficult. We show that using the Variational Information Bottleneck (Alemi et al., 2016) to regularize convolutional neural networks improves generalization to held-out domains and reduces the sim-to-real gap in a sensor-agnostic manner. As a side effect, the learned embeddings also provide useful estimates of model uncertainty for each sensor. We demonstrate that our method is able to help close the sim-to-real gap and successfully fuse RGB and depth modalities based on understanding of the situational uncertainty of each sensor. In a real-world office environment, we achieve 96% task success, improving upon the baseline by +16%.
It Takes Four to Tango: Multiagent Selfplay for Automatic Curriculum Generation

Yuqing Du, Pieter Abbeel, and Aditya Grover

International Conference on Learning Representations (ICLR) 2022

Abstract arXiv Code

We are interested in training general-purpose reinforcement learning agents that can solve a wide variety of goals. Training such agents efficiently requires automatic generation of a goal curriculum. This is challenging as it requires (a) exploring goals of increasing difficulty, while ensuring that the agent (b) is exposed to a diverse set of goals in a sample efficient manner and (c) does not catastrophically forget previously solved goals. We propose Curriculum Self Play (CuSP), an automated goal generation framework that seeks to satisfy these desiderata by virtue of a multi-player game with 4 agents. We extend the asymmetric curricula learning in PAIRED (Dennis et al., 2020) to a symmetrized game that carefully balances cooperation and competition between two off-policy student learners and two regret-maximizing teachers. CuSP additionally introduces entropic goal coverage and accounts for the non-stationary nature of the students, allowing us to automatically induce a curriculum that balances progressive exploration with anti-catastrophic exploitation. We demonstrate that our method succeeds at generating an effective curricula of goals for a range of control tasks, outperforming other methods at zero-shot test-time generalization to novel out-of-distribution goals.
Auto-Tuned Sim-to-Real Transfer

Yuqing Du*, Olivia Watkins*, Trevor Darrell, Pieter Abbeel, and Deepak Pathak

International Conference on Robotics and Automation (ICRA) 2021
(Best Cognitive Robotics Paper Finalist)

Abstract arXiv Website Code

Policies trained in simulation often fail when transferred to the real world due to the `reality gap' where the simulator is unable to accurately capture the dynamics and visual properties of the real world. Current approaches to tackle this problem, such as domain randomization, require prior knowledge and engineering to determine how much to randomize system parameters in order to learn a policy that is robust to sim-to-real transfer while also not being too conservative. We propose a method for automatically tuning simulator system parameters to match the real world using only raw RGB images of the real world without the need to define rewards or estimate state. Our key insight is to reframe the auto-tuning of parameters as a search problem where we iteratively shift the simulation system parameters to approach the real-world system parameters. We propose a Search Param Model (SPM) that, given a sequence of observations and actions and a set of system parameters, predicts whether the given parameters are higher or lower than the true parameters used to generate the observations. We evaluate our method on multiple robotic control tasks in both sim-to-sim and sim-to-real transfer, demonstrating significant improvement over naive domain randomization.
AvE: Assistance via Empowerment

Yuqing Du, Stas Tiomkin, Emre Kiciman, Daniel Polani, Pieter Abbeel, and Anca Dragan

Neural Information Processing Systems (NeurIPS) 2020

Abstract arXiv Website Code

One difficulty in using artificial agents for human-assistive applications lies in the challenge of accurately assisting with a person’s goal(s). Existing methods tend to rely on inferring the human’s goal, which is challenging when there are many potential goals or when the set of candidate goals is difficult to identify. We propose a new paradigm for assistance by instead increasing the human’s ability to control their environment, and formalize this approach by augmenting reinforcement learning with human empowerment. This task-agnostic objective preserves the person’s autonomy and ability to achieve any eventual state. We test our approach against assistance based on goal inference, highlighting scenarios where our method overcomes failure modes stemming from goal ambiguity or misspecification. As existing methods for estimating empowerment in continuous domains are computationally hard, precluding its use in real time learned assistance, we also propose an efficient empowerment-inspired proxy metric. Using this, we are able to successfully demonstrate our method in a shared autonomy user study for a challenging simulated teleoperation task with human-in-the-loop training.
Robust Reinforcement Learning using Adversarial Populations

Eugene Vinitsky*, Yuqing Du*, Kanaad Parvate*, Kathy Jang, Pieter Abbeel, and Alexandre Bayen

Abstract arXiv Code

Reinforcement Learning (RL) is an effective tool for controller design but can struggle with issues of robustness, failing catastrophically when the underlying system dynamics are perturbed. The Robust RL formulation tackles this by adding worst-case adversarial noise to the dynamics and constructing the noise distribution as the solution to a zero-sum minimax game. However, existing work on learning solutions to the Robust RL formulation has primarily focused on training a single RL agent against a single adversary. In this work, we demonstrate that using a single adversary does not consistently yield robustness to dynamics variations under standard parametrizations of the adversary; the resulting policy is highly exploitable by new adversaries. We propose a population-based augmentation to the Robust RL formulation in which we randomly initialize a population of adversaries and sample from the population uniformly during training. We empirically validate across robotics benchmarks that the use of an adversarial population results in a more robust policy that also improves out-of-distribution generalization. Finally, we demonstrate that this approach provides comparable robustness and generalization as domain randomization on these benchmarks while avoiding a ubiquitous domain randomization failure mode.
Group Surfing: A Pedestrian-Based Approach to Sidewalk Robot Navigation

Yuqing Du, Nicholas J. Hetherington , Chu Lip Oon, Wesley P. Chan, Camilo P. Quintero, Elizabeth Croft, and Machiel H. F. Van der Loos

International Conference on Robotics and Automation (ICRA) 2019

Abstract arXiv

In this paper, we propose a novel navigation system for mobile robots in pedestrian-rich sidewalk environments. Sidewalks are unique in that the pedestrian-shared space has characteristics of both roads and indoor spaces. Like vehicles on roads, pedestrian movement often manifests as linear flows in opposing directions. On the other hand, pedestrians also form crowds and can exhibit much more random movements than vehicles. Classical algorithms are insufficient for safe navigation around pedestrians and remaining on the sidewalk space. Thus, our approach takes advantage of natural human motion to allow a robot to adapt to sidewalk navigation in a safe and socially-compliant manner. We developed a group surfing method which aims to imitate the optimal pedestrian group for bringing the robot closer to its goal. For pedestrian-sparse environments, we propose a sidewalk edge detection and following method. Underlying these two navigation methods, the collision avoidance scheme is human-aware. The integrated navigation stack is evaluated and demonstrated in simulation. A hardware demonstration is also presented.