Singapore

We propose Multi-Agent Reflective Policy Optimization (MARPO) to alleviate the issue of sample inefficiency in multi-agent reinforcement learning. MARPO introduces the reflection mechanism into multi-agent settings, effectively leveraging subsequent trajectories to improve sample efficiency. We theoretically derive an asymmetric clipping mechanism that dynamically adjusts the clipping range based on the KL divergence to overcome the limitations of fixed clipping boundaries and improve the stability of the training process. We evaluate MARPO on the StarCraft II Multi-Agent Challenge (SMAC) benchmark, including both standard SMAC tasks and the more challenging SMAC-Hard variants, with results demonstrating its superior performance.The code is provided in the supplementary material.

AAAI 2026 Main Conference

MARPO: A Reflective Policy Optimization for Multi-Agent Reinforcement Learning

poster

We are pleased to announce the Fortieth AAAI Conference on Artificial Intelligence (AAAI-26), which will be held in Singapore EXPO from January 20 to January 27, 2026.

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-26 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.<br><br>

To access this event page, you need to log in with the **email address you registered with**. <br>Access credentials will be sent to your email from Underline -  subject line "Welcome to AAAI 2026". Please be sure to check your spam email folder if you do not see an email confirmation right away.

Please log in

To access this event page, you are required to register.
Please complete your registration to continue.

We recommend reading [**the registration information**](https://aaai.org/conference/aaai/aaai-26/registration/) first.

**Online Registration Form**: https://aaai.getregistered.net/conference-2026 

Registration Required

We are pleased to announce the Fortieth AAAI Conference on Artificial Intelligence (AAAI-26), which will be held in Singapore EXPO from January 20 to January 27, 2026.

Large language models (LLMs) have demonstrated their instruction-following capabilities and achieved powerful performance on various tasks. Inspired by their success, recent works in the molecular domain have led to the development of large molecular language models (LMLMs) that integrate 1D molecular strings or 2D molecular graphs into the language models. However, existing LMLMs often suffer from hallucination and limited robustness, largely due to inadequate integration of diverse molecular modalities such as 1D sequences, 2D molecular graphs, and 3D conformations. To address these limitations, we propose CoLLaMo, a large language model-based molecular assistant equipped with a multi-level molecular modality-collaborative projector. The relation-aware modality-collaborative attention mechanism in the projector facilitates fine-grained and relation-guided information exchange between atoms by incorporating 2D structural and 3D spatial relations. Furthermore, we present a molecule-centric new automatic measurement, including a hallucination assessment metric and GPT-based caption quality evaluation to address the limitations of token-based generic evaluation metrics (i.e., BLEU) widely used in assessing molecular comprehension of LMLMs. Our extensive experiments demonstrate that our CoLLaMo enhances the molecular modality generalization capabilities of LMLMs, achieving the best performance on multiple tasks, including molecule captioning, computed property QA, descriptive property QA, motif counting, and IUPAC name prediction.

Improving Large Molecular Language Model via Relation-aware Multimodal Collaboration

In recent years, neural image compression methods have achieved impressive performance in image compression tasks, most of which are based on variational auto-encoder with hyper-prior and autoregressive Gaussian entropy model. We first demonstrate that the way these end-to-end approaches handle quantization during training leads to a mismatch between the gradients direction of entropy model parameters (i.e., mean and standard deviation) and the direction they should be optimized towards during inference, making neural network difficult to learn accurate estimates of entropy model parameters. To address this issue, we then propose a two-step improvement: in the first step, use straight-through estimator to align the forward propagation during training with inference, thereby correcting the gradients of standard deviation parameters; in the second step, utilize gradients transfer that we propose and MSE-guided gradients to manually compensate for the gradients of mean parameters lost due to straight-through estimator. Finally, we also propose to freeze the auto-encoder and hyper auto-encoder in pre-trained models provided by existing works, and fine-tune only the modules that predict the entropy model parameters, enabling efficient validation of proposed improvements. Experimental results show that our improvements bring appreciable performance gains to state-of-the-art neural image compression models in recent years. Meanwhile, our improvements require no modification to the structure of pre-trained models and only lightweight fine-tuning, which shows strong plug-and-play capability and practical utility.

Correcting Quantization-Induced Gradient Mismatch in Neural Image Compression

Multi-trait Essay Scoring (MES) aims to evaluate the quality of essays across multiple traits (e.g., Language, Content, and Organization). The task can be summarized into three crucial steps: essay content encoding, trait feature learning, and multi-trait scoring. However, previous methods fall short in these steps due to neglecting essential scoring-oriented knowledge, leading to suboptimal performance. To track these issues, we propose a novel multi-trait scoring framework with multi-knowledge enhancement. Specifically, linguistic knowledge is used to model syntactic structural relations between words, highlighting structurally-informed essay encoding. We learn trait knowledge by capturing the knowledge dependencies between traits to enhance trait-specific features. Further, score-aware ordinal knowledge is integrated to promote ordinal alignment in trait-specific features associated with score rankings, improving scoring performance. Extensive experiments show that our proposed method achieves significant performance.

Multi-knowledge Enhanced Graph Neural Network for Multi-trait Essay Scoring

Time series forecasting faces a fundamental challenge: the uneven distribution of predictive importance in time series data, where some specific time points and feature combinations carry disproportionately predictive power. As a result, uniform processing methods that treat all data alike inevitably fall short of optimal performance. To address this problem, we propose FeTS, a feature-aware framework that comprehensively learns temporal features through two key components: (i) Adaptive Feature Extraction (AdaFE), which dynamically discovers the most important features within each temporal patch and extracts them on the fly, yielding sharper and more focused local representations; and (ii) Dual-Scale Feed-Forward Network (DSFFN), which strategically integrates fine-grained local features with global long-term dependencies to achieve richer dual-scale representation learning. Extensive experiments on eight benchmark datasets demonstrate that FeTS achieves state-of-the-art performance in time series forecasting tasks, offering a novel solution to the challenge of uneven predictive importance in forecasting.

FeTS: A Feature-Aware Framework for Time Series Forecasting

Source-free unsupervised domain adaptation (SF-UDA), which relies only on a pre-trained source model and unlabeled target data, has gained significant attention. Pseudo-labeling, valued for its simplicity and effectiveness, is a key approach in SF-UDA. However, existing methods neglect the consistency priors of anatomical features across samples, leading them fail to revise of high-confidence noise in structurally inconsistent regions, ultimately manifesting as significant discrepancies in pseudo-labeled samples especially in limited source data scenarios. Motivated by this insight, we propose a novel Geometric Correspondence Constrained (GCC) pseudo-labeling framework. GCC first stratifies pseudo-labeled samples into high/low-quality subsets. It then refines low-quality samples by leveraging the anatomical features inherent in high-quality samples while injecting Gaussian perturbation to perturb high-confidence noise towards the decision boundaries. This process effectively mitigates high-confidence noise disruptive effect and preserves critical prior anatomical knowledge, making it particularly powerful for scenarios with limited source data. Experiments on cross-domain fundus image datasets demonstrate that our method achieves state-of-the-art performance.

Geometric Correspondence Constrained Pseudo-Label Alignment for Source-Free Domain Adaptive Fundus Image Segmentation

In many practical reinforcement learning tasks, feedback is only provided at the end of a long horizon, leading to sparse and delayed rewards. Existing reward redistribution methods typically assume that per-step rewards are independent, thus overlooking interdependencies among state–action pairs. In this paper, we propose a Gaussian Process-based Likelihood Reward Redistribution (GP-LRR) framework that addresses this issue by modeling the reward function as a sample from a Gaussian Process (GP), which explicitly captures dependencies between state–action pairs through the kernel function. By maximizing the likelihood of the observed episodic return via a leave-one-out strategy that leverages the entire trajectory, our framework inherently introduces uncertainty regularization. Moreover, we show that the conventional mean squared error (MSE)-based reward redistribution arises as a special case of our GP-LRR framework when using a degenerate kernel without observation noise. When integrated with an off-policy algorithm such as Soft Actor-Critic, GP-LRR yields dense and informative reward signals, resulting in superior sample efficiency and policy performance on several MuJoCo benchmarks.

Reward Redistribution via Gaussian Process Likelihood Estimation

Temporal Graph Neural Networks (TGNNs) are increasingly used in high-stakes domains, such as financial forecasting, recommendation systems, and fraud detection. However, their susceptibility to poisoning attacks poses a critical security risk. This work introduces **LORETTA** (**Lo**w **Re**source **T**wo-phase **T**emporal **A**ttack), a novel adversarial framework on Continuous-Time Dynamic Graphs which degrades TGNN performance by an average of **29.47%** across 4 widely used benchmark datasets and 4 State-of-the-Art (SotA) models.

LORETTA operates through a two-stage approach: (1) sparsify the graph by removing high-impact edges using any of 16 tested temporal importance metrics; (2) strategically replace removed edges with adversarial negatives via LORETTA's novel degree-preserving negative sampling algorithm. Our plug-and-play design eliminates the need for expensive surrogate models while adhering to realistic unnoticeability constraints. LORETTA degrades performance by up to **42.0%** on MOOC, **31.5%** on Wikipedia, **28.8%** on UCI, and **15.6%** on Enron. LORETTA outperforms 11 attack baselines, remains undetectable to 4 leading anomaly detection systems, and is robust to 4 SotA adversarial defense training methods, establishing its effectiveness, unnoticeability, and robustness.

LORETTA: A Low Resource Framework to Poison Continuous Time Dynamic Graphs

PIBT is a rule-based Multi-Agent Path Finding (MAPF) solver, widely used as a low-level planner or action sampler in many state-of-the-art approaches. Its primary advantage lies in its exceptional speed, enabling action selection for thousands of agents within milliseconds by considering only the immediate next timestep. However, this short-horizon design leads to poor performance in scenarios where agents have orientation and must perform time-consuming rotation actions. In this work, we present an enhanced version of PIBT that addresses this limitation by incorporating multi-action operations. We detail the modifications introduced to improve PIBT's performance while preserving its hallmark efficiency. Furthermore, we demonstrate how our method, when combined with graph guidance technique and large neighborhood search optimization, achieves state-of-the-art performance in the online LMAPF-T setting.

Enhancing PIBT via Multi-Action Operations

Small language models (SLMs) run quickly, consume little memory, and can be deployed on edge devices, making them especially appealing when compute or energy is limited. Because of these advantages, boosting SLMs' reasoning ability has become an important research goal. A common approach is to distill the long chains of thought (long-CoTs) produced by large reasoning models (LRMs) into SLMs, hoping to transfer the larger models’ strong reasoning ability. However, SLMs do not always benefit from distillation of long-CoTs. The lengthy and complex semantic steps and large amount of self-reflection contents in long-CoTs may exceed the limited learning capabilities of SLMs, and the impact of self-reflection density on the performance of SLMs is unclear. To resolve this capacity mismatch, we propose \textbf{MACoT}, a multi-agent framework that \textit{synthesizes} chains of thought (CoTs) that are more suitable for small models rather than compressing or pruning existing ones. Through the interactive collaboration among six types of agents, \textbf{MACoT} synthesizes semantically explicit, logically clear CoTs that efficiently activate a small model’s internal knowledge through a carefully designed output pattern. At the same time, the CoTs synthesized by our method can retain a small amount of self-reflection content, thereby matching the learning capability of the small model and maximizing its reasoning accuracy. We fine-tuned Qwen2.5-7B-Instruct using only 1879 synthetic CoTs, significantly improving its performance on mathematical reasoning tasks and generalizing well, outperforming models trained on 5x more data. Through experiments, we found that a modest level of self-reflection boosts small-model performance, whereas excessive reflection sharply degrades it, which shows that “teaching SLMs to think” hinges on aligning each CoT’s cognitive load with the model’s capacity.

MACoT: Synthesizing Chains of Thought for Small Models via Multi-Agent Collaboration

Visual language models (VLMs) have made significant progress in image captioning tasks, yet recent studies have found they are vulnerable to backdoor attacks. Attackers can inject undetectable perturbations into the data during inference, triggering abnormal behavior and generating malicious captions. These attacks are particularly challenging to detect and defend against due to the stealthiness and cross-modal propagation of the trigger signals. In this paper, we identify two key vulnerabilities by analyzing existing attack patterns: (1) the model exhibits abnormal attention concentration on certain regions of the input image, and (2) backdoor attacks often induce semantic drift and sentence incoherence. Based on these insights, we propose Semantic Reward Defense (SRD), a reinforcement learning framework that mitigates backdoor behavior without requiring any prior knowledge of trigger patterns. SRD learns to apply discrete perturbations to sensitive contextual regions of image inputs via a deep Q-network policy, aiming to confuse attention and disrupt the activation of malicious paths. To guide policy optimization, we design a reward signal named semantic fidelity score, which jointly assesses the semantic consistency and linguistic fluency of the generated captions, encouraging the agent to achieve a robust yet faithful output. SRD offers a trigger-agnostic, policy-interpretable defense paradigm that effectively mitigates local (TrojVLM) and global (Shadowcast) backdoor attacks, reducing ASR to 3.4% and 5.6% respectively, with less than 15% average CIDEr drop on the clean inputs.

Downloads

Next from AAAI 2026 Main Conference

Improving Large Molecular Language Model via Relation-aware Multimodal Collaboration

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from AAAI 2026 Main Conference

Improving Large Molecular Language Model via Relation-aware Multimodal Collaboration

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES

Downloads