OS-ATLAS: A Foundation Action Model for Generalist GUI Agents

Existing efforts in building GUI agents heavily rely on the availability of robust commercial Vision-Language Models (VLMs) such as GPT-4o and GeminiProVision. Practitioners are often reluctant to use open-source VLMs due to their significant performance lag compared to their closed-source counterparts, particularly in GUI grounding and Out-Of-Distribution (OOD) scenarios. To facilitate future research in this area, we developed OS-Atlas - a foundational GUI action model that excels at GUI grounding and OOD agentic tasks through innovations in both data and modeling. We have invested significant engineering effort in developing an open-source toolkit for synthesizing GUI grounding data across multiple platforms, including Windows, Linux, MacOS, Android, and the web. Leveraging this toolkit, we are releasing the largest open-source cross-platform GUI grounding corpus to date, which contains over 13 million GUI elements. This dataset, combined with innovations in model training, provides a solid foundation for OS-Atlas to understand GUI screenshots and generalize to unseen interfaces. Through extensive evaluation across six benchmarks spanning three different platforms (mobile, desktop, and web), OS-Atlas demonstrates significant performance improvements over previous state-of-the-art models. Our evaluation also uncovers valuable insights into continuously improving and scaling the agentic capabilities of open-source VLMs.

Enhancing Mathematical Reasoning in LLMs by Stepwise Correction

Best-of-N decoding methods instruct large language models (LLMs) to generate multiple solutions, score each using a scoring function, and select the highest scored as the final answer to mathematical reasoning problems. However, this repeated independent process often leads to the same mistakes, making the selected solution still incorrect. We propose a novel prompting method named Stepwise Correction (StepCo) that helps LLMs identify and revise incorrect steps in their generated reasoning paths. It iterates verification and revision phases that employ a process-supervised verifier. The verify-then-revise process not only improves answer correctness but also reduces token consumption with fewer paths needed to generate. With StepCo, a series of LLMs demonstrate exceptional performance. Notably, using GPT-4o as the backend LLM, StepCo achieves an average accuracy of 94.1 across eight datasets, significantly outperforming the state-of-the-art Best-of-N method by +2.4, while reducing token consumption by 77.8%.

SG-Nav: Online 3D Scene Graph Prompting for LLM-based Zero-shot Object Navigation

In this paper, we propose a new framework for zero-shot object navigation. Existing zero-shot object navigation methods prompt LLM with the text of spatially closed objects, which lacks enough scene context for in-depth reasoning. To better preserve the information of environment and fully exploit the reasoning ability of LLM, we propose to represent the observed scene with 3D scene graph. The scene graph encodes the relationships between objects, groups and rooms with a LLM-friendly structure, for which we design a hierarchical chain-of-thought prompt to help LLM reason the goal location according to scene context by traversing the nodes and edges. Moreover, benefit from the scene graph representation, we further design a re-perception mechanism to empower the object navigation framework with the ability to correct perception error. We conduct extensive experiments on MP3D, HM3D and RoboTHOR environments, where SG-Nav surpasses previous state-of-the-art zero-shot methods by more than 10% SR on all benchmarks, while the decision process is explainable. To the best of our knowledge, SG-Nav is the first zero-shot method that achieves even higher performance than supervised object navigation methods on the challenging MP3D benchmark.

CodeTaxo: Enhancing Taxonomy Expansion with Limited Examples via Code Language Prompts

Taxonomies play a crucial role in various applications by providing a structural representation of knowledge. The task of taxonomy expansion involves integrating emerging concepts into existing taxonomies by identifying appropriate parent concepts for these new query concepts. Previous approaches typically relied on self-supervised methods that generate annotation data from existing taxonomies. However, these methods are less effective when the existing taxonomy is small (fewer than 100 entities). In this work, we introduce \textsc{CodeTaxo}, a novel approach that leverages large language models through code language prompts to capture the taxonomic structure. Extensive experiments on five real-world benchmarks from different domains demonstrate that \textsc{CodeTaxo} consistently achieves superior performance across all evaluation metrics, significantly outperforming previous state-of-the-art methods. The code and data are available at \url{https://github.com/QingkaiZeng/CodeTaxo-Pub}.

CTE-MLO: Continuous-time and Efficient Multi-LiDAR Odometry with Localizability-aware Point Cloud Sampling

In recent years, LiDAR-based localization and mapping methods have achieved significant progress thanks to their reliable and real-time localization capability. Considering single LiDAR odometry often faces hardware failures and degradation in practical scenarios, Multi-LiDAR Odometry (MLO), as an emerging technology, is studied to enhance the performance of LiDAR-based localization and mapping systems. However, MLO can suffer from high computational complexity introduced by dense point clouds that are fused from multiple LiDARs, and the continuous-time measurement characteristic is constantly neglected by existing LiDAR odometry. This motivates us to develop a Continuous-Time and Efficient MLO, namely CTE-MLO, which can achieve accurate and real-time state estimation using multi-LiDAR measurements through a continuous-time perspective. In this paper, the Gaussian process estimation is naturally combined with the Kalman filter, which enables each LiDAR point in a point stream to query the corresponding continuous-time trajectory within its time instants. A decentralized multi-LiDAR synchronization scheme also be devised to combine points from separate LiDARs into a single point cloud without the requirement for primary LiDAR assignment. Moreover, with the aim of improving the real-time performance of MLO without sacrificing robustness, a point cloud sampling strategy is designed with the consideration of localizability. The effectiveness of the proposed method is demonstrated through various scenarios, including public datasets and real-world autonomous driving experiments. The results demonstrate that the proposed CTE-MLO can achieve high-accuracy continuous-time state estimations in real-time and is demonstratively competitive compared to other state-of-the-art methods.

Embodied Instruction Following in Unknown Environments

Enabling embodied agents to complete complex human instructions from natural language is crucial to autonomous systems in household services. Conventional methods can only accomplish human instructions in the known environment where all interactive objects are provided to the embodied agent, and directly deploying the existing approaches for the unknown environment usually generates infeasible plans that manipulate non-existing objects. On the contrary, we propose an embodied instruction following (EIF) method for complex tasks in the unknown environment, where the agent efficiently explores the unknown environment to generate feasible plans with existing objects to accomplish abstract instructions. Specifically, we build a hierarchical embodied instruction following framework including the high-level task planner and the low-level exploration controller with multimodal large language models. We then construct a semantic representation map of the scene with dynamic region attention to demonstrate the known visual clues, where the goal of task planning and scene exploration is aligned for human instruction. For the task planner, we generate the feasible step-by-step plans for human goal accomplishment according to the task completion process and the known visual clues. For the exploration controller, the optimal navigation or object interaction policy is predicted based on the generated step-wise plans and the known visual clues. The experimental results demonstrate that our method can achieve 45.09% success rate in 204 complex human instructions such as making breakfast and tidying rooms in large house-level scenes.

Large Language Models Can Self-Correct with Minimal Effort

Intrinsic self-correct was a method that instructed large language models (LLMs) to verify and correct their responses without external feedback. Unfortunately, the study concluded that the LLMs could not self-correct reasoning yet. We find that a simple yet effective verification method can unleash inherent capabilities of the LLMs. That is to mask a key condition in the question, add the current response to construct a verification question, and predict the condition to verify the response. The condition can be an entity in an open-domain question or a numeric value in a math question, which requires minimal effort (via prompting) to identify. We propose an iterative verify-then-correct framework to progressively identify and correct (probably) false responses, named ProCo. We conduct experiments on three reasoning tasks. On average, ProCo, with GPT-3.5-Turbo as the backend LLM, yields $+6.8$ exact match on four open-domain question answering datasets, $+14.1$ accuracy on three arithmetic reasoning datasets, and $+9.6$ accuracy on a commonsense reasoning dataset, compared to Self-Correct.

Instructing Large Language Models to Identify and Ignore Irrelevant Conditions

Math word problem (MWP) solving requires generating a reasoning path based on a given problem description that often contains irrelevant conditions. Existing chain-of-thought (CoT) prompting methods elicited multi-step reasoning abilities of large language models (LLMs) to solve MWPs. However, they were seriously confused by the irrelevant conditions, resulting in low accuracy. In this paper, we propose a novel approach named I$^3$C that instructs LLMs to identify and ignore irrelevant conditions. It identifies a set of irrelevant condition candidates that have a weak semantic relevance with the question. Then it prompts LLMs to verify the irrelevant conditions. Lastly it instructs the LLMs with the verification on relevant and irrelevant conditions to avoid confusion and improve reasoning paths. Moreover, we propose to select (problem, reasoning paths) pairs as demonstrations to enhance I$^3$C with few-shot reasoning. We develop I$^3$C-Select that selects the most confusing problems based on the semantic relevance measurement. We conduct extensive experiments on eight MWP datasets. I$^3$C can be combined with any CoT prompting methods to improve the performance of solving MWPs. Notably, with GPT-3.5-Turbo and I$^3$C-Select, we achieve an accuracy of 96.0 and 94.1 on GSM-IC2-1K and GSM-ICM-1K, respectively, significantly outperforming the state-of-the-art few-shot prompting method Complex-CoT by +11.7 and +11.1. Our implementation is made publicly available at https://wzy6642.github.io/I3C.github.io/.

Decoupled Federated Learning on Long-Tailed and Non-IID data with Feature Statistics

Federated learning is designed to enhance data security and privacy, but faces challenges when dealing with heterogeneous data in long-tailed and non-IID distributions. This paper explores an overlooked scenario where tail classes are sparsely distributed over a few clients, causing the models trained with these classes to have a lower probability of being selected during client aggregation, leading to slower convergence rates and poorer model performance. To address this issue, we propose a two-stage Decoupled Federated learning framework using Feature Statistics (DFL-FS). In the first stage, the server estimates the client's class coverage distributions through masked local feature statistics clustering to select models for aggregation to accelerate convergence and enhance feature learning without privacy leakage. In the second stage, DFL-FS employs federated feature regeneration based on global feature statistics and utilizes resampling and weighted covariance to calibrate the global classifier to enhance the model's adaptability to long-tailed data distributions. We conducted experiments on CIFAR10-LT and CIFAR100-LT datasets with various long-tailed rates. The results demonstrate that our method outperforms state-of-the-art methods in both accuracy and convergence rate.

TaE: Task-aware Expandable Representation for Long Tail Class Incremental Learning

Class-incremental learning (CIL) aims to train classifiers that learn new classes without forgetting old ones. Most CIL methods focus on balanced data distribution for each task, overlooking real-world long-tailed distributions. Therefore, Long-Tailed Class-Incremental Learning (LT-CIL) has been introduced, which trains on data where head classes have more samples than tail classes. Existing methods mainly focus on preserving representative samples from previous classes to combat catastrophic forgetting. Recently, dynamic network algorithms frozen old network structures and expanded new ones, achieving significant performance. However, with the introduction of the long-tail problem, merely extending task-specific parameters can lead to miscalibrated predictions, while expanding the entire model results in an explosion of memory size. To address these issues, we introduce a novel Task-aware Expandable (TaE) framework, dynamically allocating and updating task-specific trainable parameters to learn diverse representations from each incremental task, while resisting forgetting through the majority of frozen model parameters. To further encourage the class-specific feature representation, we develop a Centroid-Enhanced (CEd) method to guide the update of these task-aware parameters. This approach is designed to adaptively minimize the distances between intra-class features while simultaneously maximizing the distances between inter-class features across all seen classes. The utility of this centroid-enhanced method extends to all "training from scratch" CIL algorithms. Extensive experiments were conducted on CIFAR-100 and ImageNet100 under different settings, which demonstrates that TaE achieves state-of-the-art performance.