United States

In multi-agent reinforcement learning (MARL), a commonly considered paradigm is centralized training with decentralized execution (CTDE). However, in this framework, decentralized execution restricts the development of coordinated policies due to the local observation limitation. In this paper, we consider the cooperation among neighboring agents during execution and formulate their interactions as a graph. Thus, we introduce a novel encoder-decoder architecture named Factor-based Multi-Agent Transformer (f-MAT) that utilizes a transformer to enable the communication between neighboring agents during both training and execution. By dividing agents into different overlapping groups and representing each group with a factor, f-MAT fulfills efficient message passing among agents through factor-based attention layers. Empirical results on networked systems such as traffic scheduling and power control demonstrate that f-MAT achieves superior performance compared to strong baselines, thereby paving the way for handling complex collaborative problems.

AAAI 2025

Towards Efficient Collaboration via Graph Modeling in Reinforcement Learning

reinforcement learning

poster

We are pleased to announce the Thirty-Ninth AAAI Conference on Artificial Intelligence (AAAI-25), which will be held in Philadelphia, Pennsylvania at the Pennsylvania Convention Center from February 25 to March 4, 2025.

The purpose of the AAAI conference series is to promote research in Artificial Intelligence (AI) and foster scientific exchange between researchers, practitioners, scientists, students, and engineers across the entirety of AI and its affiliated disciplines. AAAI-25 will feature technical paper presentations, special tracks, invited speakers, workshops, tutorials, poster sessions, senior member presentations, competitions, and exhibit programs, and a range of other activities to be announced.

### [Invited Speakers](https://aaai.org/conference/aaai/aaai-25/aaai-25-invited-speakers/)

Register [here](https://aaai.org/conference/aaai/aaai-25/registration/)

The purpose of the AAAI conference series is to promote research in Artificial Intelligence (AI) and foster scientific exchange between researchers, practitioners, scientists, students, and engineers across the entirety of AI and its affiliated disciplines. AAAI-25 will feature technical paper presentations, special tracks, invited speakers, workshops, tutorials, poster sessions, senior member presentations, competitions, and exhibit programs, and a range of other activities to be announced.



The Audio-Visual Question Answering (AVQA) task involves extracting question-related audio-visual clues from both temporal and spatial perspectives to answer questions accurately. Despite the promising performance of existing multi-modal AVQA models, thanks to large-scale pre-trained models, challenges remain in the field. Firstly, aligning audiovisual information across temporal and spatial dimensions is
difficult. Secondly, the fusion of audio-visual information is often weighted inadequately, limiting model performance. To address these issues, we propose the Audio-Visual Adaptive Fusion Network (AVAF-Net), which uses contrastive learning to align audio-visual information temporally and spatially and adaptively adjusts fusion weights based on the question. Specifically, we initially align visual and audio information temporally through a temporal-alignment contrastive loss. This is followed by an audio-visual clue-mining module
that highlights question-related cues, aligning them with the vocal region spatially using spatial alignment contrastive loss. Additionally, a question-oriented adaptive fusion module assigns different weights to audio and visual modalities based on the question content and then fuses them. The fused audio-visual cues are finally used to predict the answer. Extensive experiments on the MUSIC-AVQA dataset show that AVAF-Net surpasses all baseline models, with a maximum improvement of 15.90% in average accuracy and an average improvement of 9.80%.

Audio-Visual Adaptive Fusion Network for Question Answering Based on Contrastive Learning

Few-shot point cloud semantic segmentation aims to accurately segment "unseen" new categories in point cloud scenes using limited labeled data. However, pretraining-based methods not only introduce excessive time overhead but also overlook the local structure representation among irregular point clouds. To address these issues, we propose a pretraining-free local structure fitting network for few-shot point cloud semantic segmentation, named TaylorSeg. Specifically, inspired by Taylor series, we treat the local structure representation of irregular point clouds as a polynomial fitting problem and propose a novel local structure fitting convolution, called TaylorConv. This convolution learns the low-order basic information and high-order refined information of point clouds from explicit encoding of local geometric structures. Then, using TaylorConv as the basic component, we construct two variant of TaylorSeg: a non-parametric TaylorSeg-NN and a parametric TaylorSeg-PN. The former can achieve performance comparable to existing parametric models without pretraining. For the latter, we equip it with an Adaptive Push-Pull (APP) module to mitigate the feature distribution differences between the query set and the support set. Extensive experiments validate the effectiveness of the proposed method. Notably, under the 2-way 1-shot setting, TaylorSeg-PN achieves improvements of +2.28% and +4.37% mIoU on the S3DIS and ScanNet datasets respectively, compared to the previous state-of-the-art methods.

Taylor Series-Inspired Local Structure Fitting Network for Few-shot Point Cloud Semantic Segmentation

Adversarial robustness in the context of Unsupervised Domain Adaptation (UDA) is particularly challenging due to the lack of labels in the target domain. Pseudo labels are often used to make adversarially robust models but compromise robustness and accuracy, falling short of the performance seen in adversarial deep learning models. The main challenges in achieving robustness and accuracy include ensuring reli-
able pseudo labels and developing effective training methods that align clean and adversarial examples during model training. To address these challenges, we propose a novel train-
ing method, Dual Source Guided Adversarial Training (DS-GAT), which jointly trains on clean and adversarial target examples using scratch source and target models with pseudo labels generated from the pretrained source model. We next update the pseudo labels to ensure their reliability against the
inevitable misclassified examples in the target domain during training. We further integrate the Mart loss and minimize the Kullback-Leibler (KL) divergence between the softmax outputs of clean and adversarial examples to significantly enhance robustness and accuracy across multiple benchmark
datasets, including OfficeHome, PACS, and VisDA. Our proposed method achieves an average relative improvement in accuracy of 8.6% and in robustness of 10.2% compared to state-of-the-art methods on the PACS dataset.

Toward Improving Robustness and Accuracy in Unsupervised Domain Adaptation

Automatic Heuristic Design (AHD) is an active research area due to its utility in solving complex search and NP-hard combinatorial optimization problems in the real world. The recent advancements in Large Language Models (LLMs) introduce new possibilities by coupling LLMs with evolutionary computation to automatically generate heuristics, known as LLM-based Evolutionary Program Search (LLM-EPS). While previous LLM-EPS studies obtained great performance on various tasks, there is still a gap in understanding the properties of heuristic search spaces and achieving a balance between exploration and exploitation, which is a critical factor in large heuristic search spaces. In this study, we address this gap by proposing two diversity measurement metrics and perform an analysis on previous LLM-EPS approaches, including FunSearch, EoH, and ReEvo. Results on black-box AHD problems reveal that while EoH demonstrates higher diversity than FunSearch and ReEvo, its objective score is unstable. Conversely, ReEvo's reflection mechanism yields good objective scores but fails to optimize diversity effectively. With this finding in mind, we introduce HSEvo, an adaptive LLM-EPS framework that maintains a balance between diversity and convergence with a harmony search algorithm. Through experimentation, we find that HSEvo achieved high diversity indices and good objective scores while remaining cost-effective. These results underscore the importance of balancing exploration and exploitation and understanding heuristic search spaces in designing frameworks in LLM-EPS.

HSEvo: Elevating Automatic Heuristic Design with Diversity-Driven Harmony Search and Genetic Algorithm Using LLMs

Prompt optimization automatically optimizes prompting expressions, unlocking the full potential of LLMs in downstream tasks. However, current prompt optimization methods are costly to train and lack sufficient interpretability. This paper proposes enhancing LLMs' reasoning performance by eliciting their causal inference ability from prompting instructions to correct answers. Specifically, we introduce the Self-Causal Instruction Enhancement (SCIE) method, which enables LLMs to generate high-quality, low-quantity observational data, then estimate the causal effect based on this data, and ultimately generate instructions with the optimized causal effect. In SCIE, the instructions are treated as the treatment, and proxy features are used to process natural language, establishing causal relationships through proxy treatments between instructions and downstream tasks. Additionally, we propose applying Object-Relational (OR) principles, where the uncovered causal relationships are treated as the inheritable class across task objects, ensuring low-cost reusability. Extensive experiments demonstrate that our method effectively generates instructions that enhance reasoning performance with reduced training cost, leveraging interpretable textual features to provide clear and actionable insights.

Eliciting Causal Abilities in Large Language Models for Reasoning Tasks

The super-alignment problem of how humans can effectively supervise super-human AI has garnered increasing attention. Recent research has focused on investigating the weak-to-strong generalization (W2SG) scenario as an analogy for super-alignment. This scenario examines how a pre-trained strong model, supervised by an aligned weak model, can outperform its weak supervisor. Despite good progress, current W2SG methods face two main issues: 1) The annotation quality is limited by the knowledge scope of the weak model; 2) It is risky to position the strong model as the final corrector.

To tackle these issues, we propose a ``Strong Empowered and Aligned Weak Mastered'' (SEAM) framework for weak annotations in W2SG. This framework can leverage the vast intrinsic knowledge of the pre-trained strong model to empower the annotation and position the aligned weak model as the annotation master. Specifically, the pre-trained strong model first generates principle fast-and-frugal trees for samples to be annotated, encapsulating rich sample-related knowledge. Then, the aligned weak model picks informative nodes based on the tree's information distribution for final annotations. Experiments on six datasets for preference tasks in W2SG scenarios validate the effectiveness of our proposed method.

Strong Empowered and Aligned Weak Mastered Annotation for Weak-to-Strong Generalization

With the rising use of neural networks across various application domains, it becomes increasingly important to ensure that they do not exhibit dangerous or undesired behaviour. In light of this, several neural network robustness verification algorithms have been developed, among which methods based on Branch and Bound (BaB) constitute the current state of the art. However, these algorithms still require immense computational resources. In this work, we seek to reduce this cost by leveraging running time prediction techniques, thereby allowing for more efficient resource allocation and use. 
Towards this end, we present a novel method that dynamically predicts whether a verification instance can be solved in the remaining time budget available to the verification algorithm. We introduce features describing BaB-based verification instances and use these to construct running time, and more specifically, timeout prediction models. We leverage these models to terminate runs on instances early in the verification process that would otherwise result in a timeout. Overall, using our method, we were able to reduce the total running time by 64% on average compared to the standard verification procedure, while certifying a comparable number of instances.

Dynamic Algorithm Termination for Branch-and-Bound-based Neural Network Verification

Large language models (LLMs) have significantly advanced the field of automated code generation. However, a notable research gap exists in the evaluation of social biases that may be present in the code produced by LLMs. To solve this issue, we propose a novel fairness framework, i.e., Solar, to assess and mitigate the social biases of LLM-generated code. Specifically, Solar can automatically generate test cases for quantitatively uncovering social biases of the auto-generated code by LLMs. To quantify the severity of social biases in generated code, we develop a dataset that covers a diverse set of social problems. We applied Solar and the crafted dataset to four state-of-the-art LLMs for code generation. Our evaluation reveals severe bias in the LLM-generated code from all the subject LLMs. Furthermore, we explore several strategies for bias mitigation,  including Chain-of-Thought (CoT) prompting, combining positive role-playing with CoT prompting and iterative prompting. Our experiments show that iterative prompting can effectively reduce social bias in LLM-generated code by up to 90%. Solar is highly extensible to evaluate new social problems.

Bias Unveiled: Investigating Social Bias in LLM-Generated Code

Multi-objective preference alignment of large language models (LLMs) is critical for developing personalizable, helpful and harmless AI systems. However, optimizing model outputs in the presence of diverse objectives, while also allowing for varying the relative weights of these objectives at inference time presents a significant challenge. Existing approaches are either computationally expensive to train or do not exhibit sufficient steerability at inference time. This paper introduces the Multi-Objective Online DPO (MO-ODPO) algorithm, designed to robustly and efficiently align model behaviors with multiple, potentially conflicting human preferences. Our approach incorporates a prompt conditioning mechanism, allowing for easy adjustment of user preferences at test time while only training a single policy, ensuring adaptive and personalized model performance. Experimental results on two popular benchmarks demonstrate the efficacy of MO-ODPO in achieving Pareto-optimal performance as well as good inference-time steerability over current state-of-the-art approaches.

Robust Multi-Objective Preference Alignment with Online DPO

We present a Reinforcement Learning Platform for Adversarial Black-box untargeted and targeted attacks that allows users to select from various distortion filters to create adversarial examples. The platform uses a Reinforcement Learning agent to add minimum distortion to input images while still causing misclassification by the target model. The agent uses a novel dual-action method to explore the input image at each step to identify sensitive regions for adding distortions while removing noises that have less impact on the target model. This dual action leads to faster and more efficient convergence of the attack. The platform can also be used to measure the robustness of image classification models against specific distortion types. Also, retraining the model with adversarial samples significantly improved robustness when evaluated on benchmark datasets. The proposed platform outperforms state-of-the-art methods in terms of the average number of queries required to cause misclassification. This advances trustworthiness with a positive social impact.

Premium content

Next from AAAI 2025

Audio-Visual Adaptive Fusion Network for Question Answering Based on Contrastive Learning

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES