Yifei Zhou

Yuda Song, Yifei Zhou, Ayush Sekhari et al. · cmu

We consider a hybrid reinforcement learning setting (Hybrid RL), in which an agent has access to an offline dataset and the ability to collect experience via real-world online interaction. The framework mitigates the challenges that arise in both pure offline and online RL settings, allowing for the design of simple and highly effective algorithms, in both theory and practice. We demonstrate these advantages by adapting the classical Q learning/iteration algorithm to the hybrid setting, which we call Hybrid Q-Learning or Hy-Q. In our theoretical results, we prove that the algorithm is both computationally and statistically efficient whenever the offline dataset supports a high-quality policy and the environment has bounded bilinear rank. Notably, we require no assumptions on the coverage provided by the initial distribution, in contrast with guarantees for policy gradient/iteration methods. In our experimental results, we show that Hy-Q with neural network function approximation outperforms state-of-the-art online, offline, and hybrid RL baselines on challenging benchmarks, including Montezuma's Revenge.

Yifei Zhou, Ayush Sekhari, Yuda Song et al. · cmu

Hybrid RL is the setting where an RL agent has access to both offline data and online data by interacting with the real-world environment. In this work, we propose a new hybrid RL algorithm that combines an on-policy actor-critic method with offline data. On-policy methods such as policy gradient and natural policy gradient (NPG) have shown to be more robust to model misspecification, though sometimes it may not be as sample efficient as methods that rely on off-policy learning. On the other hand, offline methods that depend on off-policy training often require strong assumptions in theory and are less stable to train in practice. Our new approach integrates a procedure of off-policy training on the offline data into an on-policy NPG framework. We show that our approach, in theory, can obtain a best-of-both-worlds type of result -- it achieves the state-of-art theoretical guarantees of offline RL when offline RL-specific assumptions hold, while at the same time maintaining the theoretical guarantees of on-policy NPG regardless of the offline RL assumptions' validity. Experimentally, in challenging rich-observation environments, we show that our approach outperforms a state-of-the-art hybrid RL baseline which only relies on off-policy policy optimization, demonstrating the empirical benefit of combining on-policy and off-policy learning. Our code is publicly available at https://github.com/YifeiZhou02/HNPG.

Yujie Zhang, Qi Yang, Yifei Zhou et al.

The goal of objective point cloud quality assessment (PCQA) research is to develop quantitative metrics that measure point cloud quality in a perceptually consistent manner. Merging the research of cognitive science and intuition of the human visual system (HVS), in this paper, we evaluate the point cloud quality by measuring the complexity of transforming the distorted point cloud back to its reference, which in practice can be approximated by the code length of one point cloud when the other is given. For this purpose, we first make space segmentation for the reference and distorted point clouds based on a 3D Voronoi diagram to obtain a series of local patch pairs. Next, inspired by the predictive coding theory, we utilize a space-aware vector autoregressive (SA-VAR) model to encode the geometry and color channels of each reference patch with and without the distorted patch, respectively. Assuming that the residual errors follow the multi-variate Gaussian distributions, the self-complexity of the reference and transformational complexity between the reference and distorted samples are computed using covariance matrices. Additionally, the prediction terms generated by SA-VAR are introduced as one auxiliary feature to promote the final quality prediction. The effectiveness of the proposed transformational complexity based distortion metric (TCDM) is evaluated through extensive experiments conducted on five public point cloud quality assessment databases. The results demonstrate that TCDM achieves state-of-the-art (SOTA) performance, and further analysis confirms its robustness in various scenarios. The code is publicly available at https://github.com/zyj1318053/TCDM.

Yifei Zhou, Zilu Li, Abhinav Shrivastava et al.

Modern retrieval system often requires recomputing the representation of every piece of data in the gallery when updating to a better representation model. This process is known as backfilling and can be especially costly in the real world where the gallery often contains billions of samples. Recently, researchers have proposed the idea of Backward Compatible Training (BCT) where the new representation model can be trained with an auxiliary loss to make it backward compatible with the old representation. In this way, the new representation can be directly compared with the old representation, in principle avoiding the need for any backfilling. However, followup work shows that there is an inherent tradeoff where a backward compatible representation model cannot simultaneously maintain the performance of the new model itself. This paper reports our ``not-so-surprising'' finding that adding extra dimensions to the representation can help here. However, we also found that naively increasing the dimension of the representation did not work. To deal with this, we propose Backward-compatible Training with a novel Basis Transformation ($BT^2$). A basis transformation (BT) is basically a learnable set of parameters that applies an orthonormal transformation. Such a transformation possesses an important property whereby the original information contained in its input is retained in its output. We show in this paper how a BT can be utilized to add only the necessary amount of additional dimensions. We empirically verify the advantage of $BT^2$ over other state-of-the-art methods in a wide range of settings. We then further extend $BT^2$ to other challenging yet more practical settings, including significant change in model architecture (CNN to Transformers), modality change, and even a series of updates in the model architecture mimicking the evolution of deep learning models.

Yifei Zhou, Juntao Ren, Fengyu Li et al.

Advances in the field of vision-language contrastive learning have made it possible for many downstream applications to be carried out efficiently and accurately by simply taking the dot product between image and text representations. One of the most representative approaches proposed recently known as CLIP has garnered widespread adoption due to its effectiveness. CLIP is trained with an InfoNCE loss that takes into account both positive and negative samples to help learn a much more robust representation space. This paper reveals that the common downstream practice of taking a dot product is only a zeroth-order approximation of the optimization goal, resulting in a loss of information during test-time. Intuitively, since the model has been optimized based on the InfoNCE loss, test-time procedures should also be in alignment. The question lies in how one can retrieve any semblance of negative samples information during inference in a computationally efficient way. To this end, we propose Distribution Normalization (DN), where we approximate the mean representation of a batch of test samples and use such a mean to represent what would be analogous to negative samples in the InfoNCE loss. DN requires no retraining or fine-tuning and can be effortlessly applied during inference. Extensive experiments on a wide variety of downstream tasks exhibit a clear advantage of DN over the dot product on top of other existing test-time augmentation methods.

Grace Tang, Swetha Rajkumar, Yifei Zhou et al.

Building generalist robotic systems involves effectively endowing robots with the capabilities to handle novel objects in an open-world setting. Inspired by the advances of large pre-trained models, we propose Keypoint Affordance Learning from Imagined Environments (KALIE), which adapts pre-trained Vision Language Models (VLMs) for robotic control in a scalable manner. Instead of directly producing motor commands, KALIE controls the robot by predicting point-based affordance representations based on natural language instructions and visual observations of the scene. The VLM is trained on 2D images with affordances labeled by humans, bypassing the need for training data collected on robotic systems. Through an affordance-aware data synthesis pipeline, KALIE automatically creates massive high-quality training data based on limited example data manually collected by humans. We demonstrate that KALIE can learn to robustly solve new manipulation tasks with unseen objects given only 50 example data points. Compared to baselines using pre-trained VLMs, our approach consistently achieves superior performance.

Yifei Zhou, Yansong Feng

Recent works show that discourse analysis benefits from modeling intra- and inter-sentential levels separately, where proper representations for text units of different granularities are desired to capture both the meaning of text units and their relations to the context. In this paper, we propose to take advantage of transformers to encode contextualized representations of units of different levels to dynamically capture the information required for discourse dependency analysis on intra- and inter-sentential levels. Motivated by the observation of writing patterns commonly shared across articles, we propose a novel method that treats discourse relation identification as a sequence labelling task, which takes advantage of structural information from the context of extracted discourse trees, and substantially outperforms traditional direct-classification methods. Experiments show that our model achieves state-of-the-art results on both English and Chinese datasets.

Yifei Zhou, Renyu Li, Hayden Housen et al.

Paraphrase Identification is a fundamental task in Natural Language Processing. While much progress has been made in the field, the performance of many state-of-the-art models often suffer from distribution shift during inference time. We verify that a major source of this performance drop comes from biases introduced by negative examples. To overcome these biases, we propose in this paper to train two separate models, one that only utilizes the positive pairs and the other the negative pairs. This enables us the option of deciding how much to utilize the negative model, for which we introduce a perplexity based out-of-distribution metric that we show can effectively and automatically determine how much weight it should be given during inference. We support our findings with strong empirical results.

Yifei Zhou, Qianlan Yang, Kaixiang Lin et al.

The vision of a broadly capable and goal-directed agent, such as an Internet-browsing agent in the digital world and a household humanoid in the physical world, has rapidly advanced, thanks to the generalization capability of foundation models. Such a generalist agent needs to have a large and diverse skill repertoire, such as finding directions between two travel locations and buying specific items from the Internet. If each skill needs to be specified manually through a fixed set of human-annotated instructions, the agent's skill repertoire will necessarily be limited due to the quantity and diversity of human-annotated instructions. In this work, we address this challenge by proposing Proposer-Agent-Evaluator, an effective learning system that enables foundation model agents to autonomously discover and practice skills in the wild. At the heart of PAE is a context-aware task proposer that autonomously proposes tasks for the agent to practice with context information of the environment such as user demos or even just the name of the website itself for Internet-browsing agents. Then, the agent policy attempts those tasks with thoughts and actual grounded operations in the real world with resulting trajectories evaluated by an autonomous VLM-based success evaluator. The success evaluation serves as the reward signal for the agent to refine its policies through RL. We validate PAE on challenging vision-based web navigation, using both real-world and self-hosted websites from WebVoyager and WebArena.To the best of our knowledge, this work represents the first effective learning system to apply autonomous task proposal with RL for agents that generalizes real-world human-annotated benchmarks with SOTA performances. Our open-source checkpoints and code can be found in https://yanqval.github.io/PAE/

Yifei Zhou, Xianjun Gu, Xinyu Dai et al.

Data leakage from API responses has drawn wide attention. APIs are often not fully regulated, making them easy to abuse. One common solution is to embed watermarks into API responses for traceability. However, existing watermarking methods often require modifying database content or API response data. This forces changes to business system code, and may even disrupt normal business operations because data values are altered. In this paper, we propose an original pluggable watermarking scheme based on a watermark proxy gateway and PEMark (Position Encoding-based Watermarking). The key novelty of our approach is exploiting the inherent permutation redundancy in the ordering of JSON/XML key-value pairs -- an overlooked dimension that carries no semantic information yet provides abundant encoding capacity. First, we forward server responses to the watermark proxy gateway, a design that requires zero modification to existing business systems. Then, we embed a watermark into each API response using position encoding, which reorders keys without altering any data values. To the best of our knowledge, this is the first work to achieve distortion-free API response watermarking via position encoding over a proxy gateway. Our method does not modify any data values, so normal business operations continue seamlessly after watermark embedding. Experimental results show that our framework maintains business usability while ensuring that returned API data is traceable. Compared with current mainstream schemes, our method is robust against tampering and insertion attacks (100\% similarity), and can withstand certain levels of deletion attacks.

Junhong Shen, Hao Bai, Lunjun Zhang et al. · cmu

The current paradigm of test-time scaling relies on generating long reasoning traces ("thinking" more) before producing a response. In agent problems that require interaction, this can be done by generating thinking traces before acting in the world. However, this process does not allow agents to acquire new information from the environment or adapt their behavior over time. In this work, we propose to scale test-time interaction, an untapped dimension of test-time scaling that increases the agent's interaction horizon to enable running rich behaviors such as exploration, backtracking, and dynamic re-planning within a single rollout. To demonstrate the promise of this scaling dimension, we study the domain of web agents. We first show that even prompting-based interaction scaling without any training can improve task success on web benchmarks non-trivially. Building on this, we introduce TTI (Test-Time Interaction), a curriculum-based online reinforcement learning (RL) approach that trains agents by adaptively adjusting their rollout lengths. Using a Gemma 3 12B model, TTI produces state-of-the-art open-source, open-data web agents on WebVoyager and WebArena benchmarks. We further show that TTI enables agents to balance exploration and exploitation adaptively. Our results establish interaction scaling as a powerful, complementary axis to scaling per-step compute, offering new avenues for training adaptive agents.

Hao Bai, Yifei Zhou, Li Erran Li et al.

While a number of existing approaches for building foundation model agents rely on prompting or fine-tuning with human demonstrations, it is not sufficient in dynamic environments (e.g., mobile device control). On-policy reinforcement learning (RL) should address these limitations, but collecting actual rollouts in an environment is often undesirable in truly open-ended agentic problems such as mobile device control or interacting with humans, where each unit of interaction is associated with a cost. In such scenarios, a method for policy learning that can utilize off-policy experience by learning a trained action-value function is much more effective. In this paper, we develop an approach, called Digi-Q, to train VLM-based action-value Q-functions which are then used to extract the agent policy. We study our approach in the mobile device control setting. Digi-Q trains the Q-function using offline temporal-difference (TD) learning, on top of frozen, intermediate-layer features of a VLM. Compared to fine-tuning the whole VLM, this approach saves us compute and enhances scalability. To make the VLM features amenable for representing the Q-function, we need to employ an initial phase of fine-tuning to amplify coverage over actionable information needed for value function. Once trained, we use this Q-function via a Best-of-N policy extraction operator that imitates the best action out of multiple candidate actions from the current policy as ranked by the value function, enabling policy improvement without environment interaction. Digi-Q outperforms several prior methods on user-scale device control tasks in Android-in-the-Wild, attaining 21.2% improvement over prior best-performing method. In some cases, our Digi-Q approach already matches state-of-the-art RL methods that require interaction. The project is open-sourced at https://github.com/DigiRL-agent/digiq

Yuyang Xiao, Yifei Zhou, Haoran Wang et al.

The lack of sufficiently diverse data, coupled with limited data efficiency, remains a major bottleneck for generalist robotic models, yet systematic strategies for collecting and curating such data are not fully explored. Task diversity arises from implicit factors that are sparsely distributed across multiple dimensions and are difficult to define explicitly. To address this challenge, we propose F-ACIL, a heuristic factor-aware compositional iterative learning framework that enables structured data factorization and promotes compositional generalization. F-ACIL decomposes the data distribution into structured factor spaces such as object, action, and environment. Based on the factorized formulation, we develop a factor-wise data collection and an iterative training paradigm that promotes compositional generalization over the high-dimensional factor space, leading to more effective utilization of real-world robotic demonstrations. With extensive real-world experiments, we show that F-ACIL can achieve more than 45% performance gains with 5-10$\times$ fewer demonstrations comparing to that of which without the strategy. The results suggest that structured factorization offers a practical pathway toward efficient compositional generalization in real-world robotic learning. We believe F-ACIL can inspire more systematic research on building generalizable robotic data flywheel strategies. More demonstrations can be found at: https://f-acil.github.io/

Yuexiang Zhai, Hao Bai, Zipeng Lin et al. · berkeley

Large vision-language models (VLMs) fine-tuned on specialized visual instruction-following data have exhibited impressive language reasoning capabilities across various scenarios. However, this fine-tuning paradigm may not be able to efficiently learn optimal decision-making agents in multi-step goal-directed tasks from interactive environments. To address this challenge, we propose an algorithmic framework that fine-tunes VLMs with reinforcement learning (RL). Specifically, our framework provides a task description and then prompts the VLM to generate chain-of-thought (CoT) reasoning, enabling the VLM to efficiently explore intermediate reasoning steps that lead to the final text-based action. Next, the open-ended text output is parsed into an executable action to interact with the environment to obtain goal-directed task rewards. Finally, our framework uses these task rewards to fine-tune the entire VLM with RL. Empirically, we demonstrate that our proposed framework enhances the decision-making capabilities of VLM agents across various tasks, enabling 7b models to outperform commercial models such as GPT4-V or Gemini. Furthermore, we find that CoT reasoning is a crucial component for performance improvement, as removing the CoT reasoning results in a significant decrease in the overall performance of our method.

Yifei Zhou, Andrea Zanette, Jiayi Pan et al. · berkeley

A broad use case of large language models (LLMs) is in goal-directed decision-making tasks (or "agent" tasks), where an LLM needs to not just generate completions for a given prompt, but rather make intelligent decisions over a multi-turn interaction to accomplish a task (e.g., when interacting with the web, using tools, or providing customer support). Reinforcement learning (RL) provides a general paradigm to address such agent tasks, but current RL methods for LLMs largely focus on optimizing single-turn rewards. By construction, most single-turn RL methods cannot endow LLMs with the ability to intelligently seek information over multiple turns, perform credit assignment, or reason about their past actions -- all of which are critical in agent tasks. This raises the question: how can we design effective and efficient multi-turn RL algorithms for LLMs? In this paper, we develop a framework for building multi-turn RL algorithms for fine-tuning LLMs, that preserves the flexibility of existing single-turn RL methods for LLMs (e.g., proximal policy optimization), while accommodating multiple turns, long horizons, and delayed rewards effectively. To do this, our framework adopts a hierarchical RL approach and runs two RL algorithms in parallel: a high-level off-policy value-based RL algorithm to aggregate reward over utterances, and a low-level RL algorithm that utilizes this high-level value function to train a token policy within each utterance or turn. Our hierarchical framework, Actor-Critic Framework with a Hierarchical Structure (ArCHer), can also give rise to other RL methods. Empirically, we find that ArCHer significantly improves efficiency and performance on agent tasks, attaining a sample efficiency of about 100x over existing methods, while also improving with larger model capacity (upto the 7 billion scale that we tested on).

Lingkai Kong, Haorui Wang, Wenhao Mu et al.

Aligning large language models (LLMs) with human objectives is crucial for real-world applications. However, fine-tuning LLMs for alignment often suffers from unstable training and requires substantial computing resources. Test-time alignment techniques, such as prompting and guided decoding, do not modify the underlying model, and their performance remains dependent on the original model's capabilities. To address these challenges, we propose aligning LLMs through representation editing. The core of our method is to view a pre-trained autoregressive LLM as a discrete-time stochastic dynamical system. To achieve alignment for specific objectives, we introduce external control signals into the state space of this language dynamical system. We train a value function directly on the hidden states according to the Bellman equation, enabling gradient-based optimization to obtain the optimal control signals at test time. Our experiments demonstrate that our method outperforms existing test-time alignment techniques while requiring significantly fewer resources compared to fine-tuning methods. Our code is available at https://github.com/Lingkai-Kong/RE-Control.

Jiayi Pan, Yichi Zhang, Nicholas Tomlin et al. · berkeley

We show that domain-general automatic evaluators can significantly improve the performance of agents for web navigation and device control. We experiment with multiple evaluation models that trade off between inference cost, modularity of design, and accuracy. We validate the performance of these models in several popular benchmarks for digital agents, finding between 74.4 and 92.9% agreement with oracle evaluation metrics. Finally, we use these evaluators to improve the performance of existing agents via fine-tuning and inference-time guidance. Without any additional supervision, we improve state-of-the-art performance by 29% on the popular benchmark WebArena, and achieve around 75% relative improvement in device control settings.

Yifei Zhou, Song Jiang, Yuandong Tian et al. · meta-ai

Large language model (LLM) agents need to perform multi-turn interactions in real-world tasks. However, existing multi-turn RL algorithms for optimizing LLM agents fail to perform effective credit assignment over multiple turns while leveraging the generalization capabilities of LLMs and it remains unclear how to develop such algorithms. To study this, we first introduce a new benchmark, ColBench, where an LLM agent interacts with a human collaborator over multiple turns to solve realistic tasks in backend programming and frontend design. Building on this benchmark, we propose a novel RL algorithm, SWEET-RL (RL with Step-WisE Evaluation from Training-time information), that uses a carefully designed optimization objective to train a critic model with access to additional training-time information. The critic provides step-level rewards for improving the policy model. Our experiments demonstrate that SWEET-RL achieves a 6% absolute improvement in success and win rates on ColBench compared to other state-of-the-art multi-turn RL algorithms, enabling Llama-3.1-8B to match or exceed the performance of GPT4-o in realistic collaborative content creation.

Jiayi Pan, Xiuyu Li, Long Lian et al. · berkeley

Scaling inference-time computation has substantially improved the reasoning capabilities of language models. However, existing methods have significant limitations: serialized chain-of-thought approaches generate overly long outputs, leading to increased latency and exhausted context windows, while parallel methods such as self-consistency suffer from insufficient coordination, resulting in redundant computations and limited performance gains. To address these shortcomings, we propose Adaptive Parallel Reasoning (APR), a novel reasoning framework that enables language models to orchestrate both serialized and parallel computations end-to-end. APR generalizes existing reasoning methods by enabling adaptive multi-threaded inference using spawn() and join() operations. A key innovation is our end-to-end reinforcement learning strategy, optimizing both parent and child inference threads to enhance task success rate without requiring predefined reasoning structures. Experiments on the Countdown reasoning task demonstrate significant benefits of APR: (1) higher performance within the same context window (83.4% vs. 60.0% at 4k context); (2) superior scalability with increased computation (80.1% vs. 66.6% at 20k total tokens); (3) improved accuracy at equivalent latency (75.2% vs. 57.3% at approximately 5,000ms). APR represents a step towards enabling language models to autonomously optimize their reasoning processes through adaptive allocation of computation.

Yifei Zhou, Sergey Levine, Jason Weston et al.

Large language models are quickly becoming the foundation for intelligent agents that are capable of using tools. However, training such agents is challenging because it requires human creation and annotation of a diverse set of tasks, tools, and evaluation criteria. In this paper, we propose the Self-Challenging framework for training an agent on high-quality tasks that are generated by itself. The agent first plays the role of challenger and generates a task after interacting with the given tools. The tasks take the form of a novel general class of problems termed Code-as-Task, which are defined by an instruction, a verification function and solution and failure cases which serve as tests, allowing to filter only for high-quality tasks. The agent then takes an executor role and trains on those tasks with reinforcement learning using the evaluation feedback as a reward. Evaluation on two existing multi-turn tool-use agent benchmarks, M3ToolEval and TauBench, shows the Self-Challenging framework achieves over a two-fold improvement in Llama-3.1-8B-Instruct, despite using only self-generated training data.

Sayash Kapoor, Benedikt Stroebl, Peter Kirgis et al. · microsoft-research, princeton

AI agents have been developed for complex real-world tasks from coding to customer service. But AI agent evaluations suffer from many challenges that undermine our understanding of how well agents really work. We introduce the Holistic Agent Leaderboard (HAL) to address these challenges. We make three main contributions. First, we provide a standardized evaluation harness that orchestrates parallel evaluations across hundreds of VMs, reducing evaluation time from weeks to hours while eliminating common implementation bugs. Second, we conduct three-dimensional analysis spanning models, scaffolds, and benchmarks. We validate the harness by conducting 21,730 agent rollouts across 9 models and 9 benchmarks in coding, web navigation, science, and customer service with a total cost of about $40,000. Our analysis reveals surprising insights, such as higher reasoning effort reducing accuracy in the majority of runs. Third, we use LLM-aided log inspection to uncover previously unreported behaviors, such as searching for the benchmark on HuggingFace instead of solving a task, or misusing credit cards in flight booking tasks. We share all agent logs, comprising 2.5B tokens of language model calls, to incentivize further research into agent behavior. By standardizing how the field evaluates agents and addressing common pitfalls in agent evaluation, we hope to shift the focus from agents that ace benchmarks to agents that work reliably in the real world.

Yifei Zhou, Xuchu Huang, Chenyu Ni et al.

Neuro-symbolic artificial intelligence (neuro-symbolic AI) excels in logical analysis and reasoning. Hyperdimensional Computing (HDC), a promising brain-inspired computational model, is integral to neuro-symbolic AI. Various HDC models have been proposed to represent class-instance and class-class relations, but when representing the more complex class-subclass relation, where multiple objects associate different levels of classes and subclasses, they face challenges for factorization, a crucial task for neuro-symbolic AI systems. In this article, we propose FactorHD, a novel HDC model capable of representing and factorizing the complex class-subclass relation efficiently. FactorHD features a symbolic encoding method that embeds an extra memorization clause, preserving more information for multiple objects. In addition, it employs an efficient factorization algorithm that selectively eliminates redundant classes by identifying the memorization clause of the target class. Such model significantly enhances computing efficiency and accuracy in representing and factorizing multiple objects with class-subclass relation, overcoming limitations of existing HDC models such as "superposition catastrophe" and "the problem of 2". Evaluations show that FactorHD achieves approximately 5667x speedup at a representation size of 10^9 compared to existing HDC models. When integrated with the ResNet-18 neural network, FactorHD achieves 92.48% factorization accuracy on the Cifar-10 dataset.

Chang Peng, Yifei Zhou, Huifeng Xi et al.

Marker-free human pose estimation (HPE) has found increasing applications in various fields. Current HPE suffers from occasional errors in keypoint recognition and random fluctuation in keypoint trajectories when analyzing kinematic human poses. The performance of existing deep learning-based models for HPE refinement is considerably limited by inaccurate training datasets in which the keypoints are manually annotated. This paper proposed a novel method to overcome the difficulty through joint angle-based modeling. The key techniques include: (i) A joint angle-based model of human pose, which is robust to describe kinematic human poses; (ii) Approximating temporal variation of joint angles through high order Fourier series to get reliable "ground truth"; (iii) A bidirectional recurrent network is designed as a post-processing module to refine the estimation of well-established HRNet. Trained with the high-quality dataset constructed using our method, the network demonstrates outstanding performance to correct wrongly recognized joints and smooth their spatiotemporal trajectories. Tests show that joint angle-based refinement (JAR) outperforms the state-of-the-art HPE refinement network in challenging cases like figure skating and breaking.

Hao Bai, Yifei Zhou, Mert Cemri et al.

Training corpuses for vision language models (VLMs) typically lack sufficient amounts of decision-centric data. This renders off-the-shelf VLMs sub-optimal for decision-making tasks such as in-the-wild device control through graphical user interfaces (GUIs). While training with static demonstrations has shown some promise, we show that such methods fall short for controlling real GUIs due to their failure to deal with real-world stochasticity and non-stationarity not captured in static observational data. This paper introduces a novel autonomous RL approach, called DigiRL, for training in-the-wild device control agents through fine-tuning a pre-trained VLM in two stages: offline RL to initialize the model, followed by offline-to-online RL. To do this, we build a scalable and parallelizable Android learning environment equipped with a VLM-based evaluator and develop a simple yet effective RL approach for learning in this domain. Our approach runs advantage-weighted RL with advantage estimators enhanced to account for stochasticity along with an automatic curriculum for deriving maximal learning signal. We demonstrate the effectiveness of DigiRL using the Android-in-the-Wild (AitW) dataset, where our 1.3B VLM trained with RL achieves a 49.5% absolute improvement -- from 17.7 to 67.2% success rate -- over supervised fine-tuning with static human demonstration data. These results significantly surpass not only the prior best agents, including AppAgent with GPT-4V (8.3% success rate) and the 17B CogAgent trained with AitW data (38.5%), but also the prior best autonomous RL approach based on filtered behavior cloning (57.8%), thereby establishing a new state-of-the-art for digital agents for in-the-wild device control.