Singapore

Balancing exploration and exploitation is a central goal in reinforcement learning (RL). Despite recent advances in enhancing language model (LM) reasoning, most methods lean toward exploitation, and increasingly encounter performance plateaus. In this work, we revisit entropy -- a signal of exploration in RL -- and examine its relationship to exploratory reasoning in LMs. Through empirical analysis, we uncover positive correlations between high-entropy regions and three types of exploratory reasoning actions: (1) pivotal tokens that determine or connect logical steps, (2) reflective actions such as self-verification and correction, and (3) rare behaviors under-explored by the base LMs. Motivated by this, we introduce a minimal modification to standard RL with only one line of code: augmenting the advantage function with an entropy-based term. Unlike traditional maximum-entropy methods which encourage exploration by promoting uncertainty, we encourage exploration by promoting deeper and longer reasoning chains. Notably, our method achieves significant gains on the Pass@K metric -- an upper-bound estimator of LM reasoning capabilities -- even when evaluated with extremely large K values, pushing the boundaries of LM reasoning.

AAAI 2026

Reasoning with Exploration: An Entropy Perspective

exploration in reinforcement learning

large language model

reasoning

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.

poster

Brain-assisted target speaker extraction (TSE) isolates a target speaker's voice from a mixture by leveraging task-specific representations in Electroencephalogram (EEG) signals. However, existing methods rely on fixed interpolation for EEG-audio alignment, introducing redundant computations. They also employ single-path encoders that extract only target-relevant features while neglecting complementary, irrelevant ones, limiting discriminability. To address these limitations, this paper proposes a $\textbf{T}$rainable EEG $\textbf{I}$nterpolation and Structure-sharing $\textbf{D}$ual-path $\textbf{E}$ncoders network (TIDENet). The proposed Trainable EEG Interpolation (TEI) uses a neural network module to leverage cross-sample EEG information during resampling by parameters updating, thereby overcoming the limitations of fixed interpolation. The Structure-sharing Dual-path Encoders (SSDPE) extend existing speech and EEG encoders by introducing dual paths that separately process features relevant and irrelevant to the target speaker and incorporates interactive fusion between them, which enhances the encoder's ability to capture task-relevant information. Experimental results on public datasets demonstrate that TIDENet achieves relative improvements of up to $\textbf{20.47}$%, $\textbf{22.22}$%, $\textbf{2.91}$%, $\textbf{6.20}$%, and $\textbf{15.84}$% in signal-to-distortion ratio (SDR), scale-invariant SDR (SI-SDR), short-time objective intelligibility (STOI), extended STOI (ESTOI), and perceptual evaluation of speech quality (PESQ), respectively, compared to the state-of-the-art. These significant gains validate the effectiveness of the proposed TEI method and SSDPE architecture.

Trainable EEG Interpolation and Structure-Sharing Dual-Path Encoders for Brain-Assisted Target Speaker Extraction

Vision-Language Models (VLMs) like CLIP struggle to understand negation, often embedding affirmatives and negatives similarly (e.g., matching "no dog" with dog images). Existing methods refine negation understanding via fine-tuning CLIP’s text encoder, risking overfitting. In this work, we propose CLIPGlasses, a plug-and-play framework that enhances CLIP’s ability to comprehend negated visual descriptions. CLIPGlasses adapts a dual-stage design: a Lens module disentangles negated semantics from text embeddings, and a Frame module predicts context-aware repulsion strength, which is integrated into the modified similarity computation to penalize alignment with negated semantics, thereby reducing false positive matches. Experiments show that CLIP equipped with CLIPGlasses achieves competitive in-domain performance and outperforms state-of-the-art methods in cross-domain generalization. Its superiority is especially evident under low-resource conditions, indicating stronger robustness across domains. Source code is included in the supplementary material.

Not Just What’s There: Enabling CLIP to Comprehend Negated Visual Descriptions Without Fine-Tuning

Deep neural networks are susceptible to adversarial examples, which induce incorrect predictions through imperceptible perturbations.
Transfer-based attacks create adversarial examples for surrogate models and transfer these examples to target models under black-box scenarios. Recent studies have established a strong correlation between the geometric properties of loss landscapes and the transferability of adversarial examples, demonstrating that flatter loss surfaces consistently yield superior transferability. However, we identify that these methods fail to account for the loss landscape flatness along the path from the current point to local minima, resulting in poor transferability. 
To address this, this paper constructs a novel Path Flatness Attack (PFA) method to significantly enhance the transferability of adversarial examples. Specifically, this paper proposes a novel path flatness indicator that not only evaluates the flatness in local minima regions but also explicitly quantifies the loss surface geometry along the trajectory from the current point to the minimum. Furthermore, we incorporate the path flatness indicator into the attack process, integrating penalties over low-loss points along the path while maximizing the loss function, thereby explicitly flattening the loss landscape. Extensive experiments demonstrate that PFA consistently achieves state-of-the-art attack performance across all experimental settings.

Prompting Adversarial Transferability via Path Flatness Attack

Federated learning (FL) has shown success in collaboratively training a model among decentralized data resources without directly sharing privacy-sensitive training data. Despite recent advances, non-IID (non-independent and identically distributed) data poses an inevitable challenge that hinders the use of FL. In this work, we address the issue of non-IID histopathological images with feature distribution shifts from an intuitive perspective that has only received limited attention. Specifically, we address this issue from the perspective of data distribution by solely adjusting the data distributions of all clients. Building on the success of diffusion models in fitting data distributions and leveraging stain separation to extract the pivotal features that are closely related to the non-IID properties of histopathological images, we propose a Federated Stain Distribution Alignment (FedSDA) method. FedSDA aligns the stain distribution of each client with a target distribution in an FL framework to mitigate distribution shifts among clients. Furthermore, considering that training
diffusion models on raw data in FL has been shown to be susceptible to privacy leakage risks, we circumvent this problem while still effectively achieving alignment. Extensive experimental results show that FedSDA is not only effective in improving baselines that focus on mitigating disparities across clients’ model updates but also outperforms baselines that address the non-IID data issues from the perspective of data distribution. We show that FedSDA provides valuable and practical insights for the computational pathology community.

FedSDA: Federated Stain Distribution Alignment for Non-IID Histopathological Image Classification

Federated learning has emerged as a promising paradigm for collaborative model training while preserving data privacy. However, many existing FL methods implicitly assume that clients have sufficient computational and storage resources, making them less applicable in real-world scenarios with severe system heterogeneity. To address this, submodel extraction has recently gained attention as a promising strategy to tailor the global model to resource-constrained clients. Despite this progress, existing methods often suffer from noticeable performance gaps across clients and structural inconsistency in the extracted models, leading to degraded global performance and increased communication overhead. In this work, we propose FedLAGC, a novel federated framework that jointly tackles performance imbalance and communication inefficiency through Layer-Adaptive submodel extraction and Gradient Correction. Specifically, FedLAGC constructs client-specific submodels by selecting structurally important parameters according to layer-wise importance scores, ensuring both resource adaptiveness and architectural consistency. Additionally, we propose a lightweight correction mechanism that captures historical optimization drift, helping to align local updates with the global direction and reduce redundant communication. The rigorous convergence analysis of FedLAGC for system-heterogeneous federated learning under non-convex objectives is given. Extensive experiments on CIFAR-10 and CIFAR-100 with ResNet-18 and ResNet-34 under various system and data heterogeneity settings demonstrate the significant superiority of FedLAGC (up to 24\% accuracy improvement and 3.66$\times$ communication efficiency) over state-of-the-art methods.

FedLAGC: Towards High Performance System-Heterogeneous Federated Learning via Layer-Adaptive Submodel Extraction and Gradient Correction

Federated learning synchronizes models through gradient transmission and aggregation. However, these gradients pose significant privacy risks, as sensitive training data is embedded within them. Existing gradient-based reconstruction attacks suffer from significantly degraded reconstruction quality when gradients are perturbed by noise-a common defense mechanism. In this paper, we introduce Gradient-Guided Conditional Diffusion Models (GG-CDMs) for reconstructing private images from leaked gradients without prior conditions. Our approach leverages the inherent denoising capabilities of diffusion models to circumvent the partial protection offered by noise perturbation, thereby enhancing attack efficacy under such defenses. Furthermore, we provide a rigorous theoretical analysis of reconstruction error bounds and the decrease rate of attack loss, characterizing the relationship between noise magnitude, model architectures, and reconstruction quality. Extensive experiments validate the effectiveness of our method and confirm our theoretical findings, demonstrating our method's superior reconstruction quality from noise-perturbed gradients by leveraging GG-CDMs.

Enhanced Privacy Leakage from Noise-Perturbed Gradients via Gradient-Guided Conditional Diffusion Models

Recently, Large Language Models (LLMs) based Web Agents have shown significant potential in web understanding and interaction tasks. However, their personalization ability and user experience remain limited by the ambiguity and dynamic nature of user intent, struggling to model diverse user interests and track intent changes over time. To address these challenges, this paper proposes Orion, a novel personalized Web Agent. Orion adopts a global-micro profiling mechanism to balance users' long-term stable preferences and scenario-based needs, and introduces context-aware interest retrieval to enhance personalization. Additionally, we design adaptive profile tracking and proactive disambiguation mechanisms to effectively address the continuous evolution of user intent in multi-turn interactions. Orion is optimized through end-to-end online reinforcement learning, improving personalized reasoning and decision-making ability in real interactive scenarios. Experiments demonstrate that Orion significantly outperforms state-of-the-art baselines in personalized understanding and task efficiency.

Orion: Steering Personalized Web Agents via Global-Micro Profiling and Adaptive Intent Tracking

Reconstructing topologically consistent facial geometry is crucial for the digital avatar creation pipelines. Existing methods either require tedious manual efforts, lack generalization to in-the-wild data, or are constrained by the limited expressiveness of 3D Morphable Models. To address these limitations, we propose VGGTFace, an automatic approach that innovatively applies the 3D foundation model, i.e. VGGT, for topologically consistent facial geometry reconstruction from in-the-wild multi-view images captured by everyday users. Our key insight is that, by leveraging VGGT, our method naturally inherits strong generalization ability and expressive power from its large-scale training and point map representation. However, it is unclear how to reconstruct a topologically consistent mesh from VGGT, as the topology information is missing in its prediction. To this end, we augment VGGT with Pixel3DMM for injecting topology information via pixel-aligned UV values. In this manner, we convert the pixel-aligned point map of VGGT to a point cloud with topology. Tailored to this point cloud with known topology, we propose a novel Topology-Aware Bundle Adjustment strategy to fuse them, where we construct a Laplacian energy for the Bundle Adjustment objective. Our method achieves high-quality reconstruction in 10 seconds for 16 views on a single NVIDIA RTX 4090. Experiments demonstrate state-of-the-art results on benchmarks and impressive generalization to in-the-wild data. Code will be released upon acceptance.

VGGTFace: Topologically Consistent Facial Geometry Reconstruction in the Wild

Recent 4D reconstruction methods have yielded impressive results but rely on sharp videos as supervision. However, motion blur often occurs in videos due to camera shake and object movement, while existing methods render blurry results when using such videos for reconstructing 4D models. Although a few approaches attempted to address the problem, they struggled to produce high-quality results, due to the inaccuracy in estimating continuous dynamic representations within the exposure time. Encouraged by recent works in 3D motion trajectory modeling using 3D Gaussian Splatting (3DGS), we take 3DGS as the scene representation manner, and propose Deblur4DGS to reconstruct a high-quality 4D model from blurry monocular video. Specifically, we transform continuous dynamic representations estimation within an exposure time into the exposure time estimation. Moreover, we introduce the exposure regularization term, multi-frame, and multi-resolution consistency regularization term to avoid trivial solutions. Furthermore, to better represent objects with large motion, we suggest blur-aware variable canonical Gaussians. Beyond novel-view synthesis, Deblur4DGS can be applied to improve blurry video from multiple perspectives, including deblurring, frame interpolation, and video stabilization. Extensive experiments in both synthetic and real-world data on the above four tasks show that Deblur4DGS outperforms state-of-the-art 4D reconstruction methods. The codes will be publicly available.

Deblur4DGS: 4D Gaussian Splatting from Blurry Monocular Video

Text-to-video generation poses significant challenges due to the inherent complexity of video data, which spans both temporal and spatial dimensions. It introduces additional redundancy, abrupt variations, and a domain gap between language and vision tokens while generation. Addressing these challenges requires an effective video tokenizer that can efficiently encode video data while preserving essential semantic and spatiotemporal information, serving as a critical bridge between text and vision. Inspired by the observation in VQ-VAE-2, we propose \textbf{HiTVideo}, a novel approach for text-to-video generation with hierarchical tokenizers. It utilizes a 3D causal VAE with a multi-layer discrete token framework, encoding video content into hierarchically structured codebooks. Higher layers capture semantic information with higher compression, while lower layers focus on fine-grained spatiotemporal details, striking a balance between compression efficiency and reconstruction quality. Our approach efficiently encodes longer video sequences (e.g., 8 seconds, 64 frames), reducing bits per pixel (bpp) by approximately 70\% compared to previous tokenizers, while maintaining competitive reconstruction quality. We explore the trade-offs between compression and reconstruction, while emphasizing the advantages of high-compressed semantic tokens in text-to-video tasks. HiTVideo aims to address the potential limitations of existing video tokenizers in text-to-video generation tasks, striving for higher compression ratios, improved token quality, and simplify LLMs modeling under language guidance, offering a scalable and promising framework for advancing text to video generation.

Content not yet available

Next from AAAI 2026

Trainable EEG Interpolation and Structure-Sharing Dual-Path Encoders for Brain-Assisted Target Speaker Extraction

Similar lecture

When Instinct Guides and Insight Grounds: Staged RL Training for LLM Agents