Singapore

Real-world text classification datasets frequently exhibit long-tail distributions, where numerous classes have sparse data, significantly degrading model performance on these underrepresented categories. While Large Language Models (LLMs) offer promise for data augmentation, existing methods often produce semantically limited samples, neglect &quot;implicit long-tails&quot; (sparse sub-patterns within classes), and lack cost-effective optimization. To address these challenges, we propose \textbf{LADA-LD (LLM-driven Adaptive Data Augmentation framework for Long-tail Distributions)}, a novel cognitive-inspired framework emulating the human learning process of &quot;recognize, explore, generate, and optimize.&quot; LADA-LD systematically enhances augmented data diversity by first detecting both explicit and implicit long-tails. It then employs an LLM for diversity-aware planning of augmentation strategies, followed by conditional generation. A low-overhead quality and diversity validator filters the synthetic data, and an adaptive incremental sampler refines future augmentation efforts based on proxy model feedback, ensuring efficient and budget-aware optimization. Extensive experiments on multiple public text classification datasets demonstrate LADA-LD&#39;s superiority over state-of-the-art methods in improving tail-class performance and overall model robustness by generating more diverse and high-fidelity augmented data. The source code is available at \url{https://anonymous.4open.science/r/DEALT-FAD0/}.

AAAI 2026

DEALT: LLM-driven Diversity-Enhanced Data Augmentation for Long-Tail Text Classification

text classification & sentiment analysis

(large) language models

data stream mining

### Welcome to ACL 2023, the 61st Annual Meeting of the Association for Computational Linguistics! 
 The conference will be held in Toronto, Canada, July 9-14, 2023. 
Following the succession of the recent conferences in our field, ACL 2023 will adopt a hybrid format.
While the impact of Covid has considerably diminished in terms of traveling, obtaining visas to Canada
entails a very long process. Moreover, the global economic conditions pose challenges for many individuals to travel to conferences. Recognizing these circumstances, we know many participants may not be
able to attend the conference in person. Therefore, we are committed to providing a great virtual platform
so everyone has the opportunity to interact with other participants and enjoy the conference. Based on the
current registered participants, approxiately 30% have chosen to attend the conference virtually. Whether
you join us in person or virtually, we sincerely hope everyone has a remarkable conference experience. 
This General Chair’s message is where I express my gratitude to the many individuals who have made
enormous contributions to the conference over the past year.

Read [**ACL 2023 General Chair's message**](https://docs.google.com/document/d/1WobYM7norbG4dI48s75HfJoD89qgX5a_F-6U8AteLSA/edit?usp=sharing/) in full.

##### **[Conference Handbook](https://2023.aclweb.org/downloads/acl2023-handbook.pdf)**

ACL 2023

The Association for Computational Linguistics (ACL) is the premier international scientific and professional society for people working on computational problems involving human language, a field often referred to as either computational linguistics or natural language processing.

Real-world text classification datasets frequently exhibit long-tail distributions, where numerous classes have sparse data, significantly degrading model performance on these underrepresented categories. While Large Language Models (LLMs) offer promise for data augmentation, existing methods often produce semantically limited samples, neglect "implicit long-tails" (sparse sub-patterns within classes), and lack cost-effective optimization. To address these challenges, we propose \textbf{LADA-LD (LLM-driven Adaptive Data Augmentation framework for Long-tail Distributions)}, a novel cognitive-inspired framework emulating the human learning process of "recognize, explore, generate, and optimize." LADA-LD systematically enhances augmented data diversity by first detecting both explicit and implicit long-tails. It then employs an LLM for diversity-aware planning of augmentation strategies, followed by conditional generation. A low-overhead quality and diversity validator filters the synthetic data, and an adaptive incremental sampler refines future augmentation efforts based on proxy model feedback, ensuring efficient and budget-aware optimization. Extensive experiments on multiple public text classification datasets demonstrate LADA-LD's superiority over state-of-the-art methods in improving tail-class performance and overall model robustness by generating more diverse and high-fidelity augmented data. The source code is available at \url{https://anonymous.4open.science/r/DEALT-FAD0/}.

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.

To access this event page, you need to log in with the **email address you registered with**. 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.

Infrared and visible image fusion aims to integrate complementary multi-modal information into a single fused result. However, existing methods 1) fail to account for the degradation visible images under adverse weather conditions, thereby compromising fusion performance; and 2) rely on fixed network architectures, limiting their adaptability to diverse degradation scenarios. To address these issues, we propose a one-stop degradation-aware image fusion framework for multi-degradation scenarios driven by a large language model (MdaIF). Given the distinct scattering characteristics of different degradation scenarios (e.g., haze, rain, and snow) in atmospheric transmission, a mixture-of-experts (MoE) system is introduced to tackle image fusion across multiple degradation scenarios. To adaptively extract diverse weather-aware degradation knowledge and scene feature representations, collectively referred to as the semantic prior, we employ a pre-trained vision-language model (VLM) in our framework. Guided by the semantic prior, we propose degradation-aware channel attention module (DCAM), which employ degradation prototype decomposition to facilitate multi-modal feature interaction in channel domain. In addition, to achieve effective expert routing, the semantic prior and channel-domain modulated features are utilized to guide the MoE, enabling robust image fusion in complex degradation scenarios. Extensive experiments validate the effectiveness of our MdaIF, demonstrating superior performance over SOTA methods. The code will be released in the future.

MdaIF: Robust One-Stop Multi-Degradation-Aware Image Fusion with Language-Driven Semantics

Recent advances in Deep Learning (DL) have improved multivariate time series (MTS) classification and regression by capturing complex patterns, but their lack of transparency hinders decision-making. Explainable AI (XAI) methods offer partial insights, yet often fall short of conveying the full decision space. Counterfactual Explanations (CE) provide a promising alternative, but current approaches typically prioritize either accuracy or sparsity -- rarely both -- limiting their practical value. To address this, we propose CONFETTI, a novel multi-objective CE method for MTS. CONFETTI uses Class Activation Maps (CAMs) to identify key subsequences, locates a counterfactual target, and optimally modifies the time series to balance prediction confidence and sparsity. This method provides actionable insights with minimal changes, improving interpretability, and decision support. CONFETTI is evaluated on seven MTS datasets from the UEA archive, demonstrating its effectiveness in various domains. CONFETTI consistently outperforms state-of-the-art CE methods in its optimization objectives, and in six other metrics from the literature, achieving, for example, at least 10% higher confidence while improving sparsity in at least 40%.

Counterfactual eXplainable AI (XAI) Method for Deep Learning-Based Multivariate Time Series Classification

Retrieval-Augmented Generation (RAG) systems enhance Large Language Models (LLMs) by retrieving relevant documents from external corpora before generating responses. This approach significantly expands LLM capabilities by leveraging vast, up-to-date external knowledge. However, this reliance on external knowledge makes RAG systems vulnerable to corpus poisoning attacks that manipulate generated outputs via poisoned document injection. Existing poisoning attack strategies typically treat the retrieval and generation stages as disjointed, limiting their effectiveness. We propose Joint-GCG, the first framework to unify gradient-based attacks across both retriever and generator models through three innovations: (1) Cross-Vocabulary Projection for aligning embedding spaces, (2) Gradient Tokenization Alignment for synchronizing token-level gradient signals, and (3) Adaptive Weighted Fusion for dynamically balancing attacking objectives. Evaluations demonstrate that Joint-GCG achieves at most 25% and an average of 5% higher attack success rate than previous methods across multiple retrievers and generators. While optimized under a white-box assumption, the generated poisons show unprecedented transferability to unseen models. Joint-GCG's innovative unification of gradient-based attacks across retrieval and generation stages fundamentally reshapes our understanding of vulnerabilities within RAG systems.

Joint-GCG: Unified Gradient-Based Poisoning Attacks on Retrieval-Augmented Generation Systems

Certified defenses aim to provide provable robustness against attacks. Models certified against adversarial attacks provide $l_p$-bounded guarantees on the absence of adversarial manipulation of inputs for predicted labels. We study the potential for malicious exploitation of certification frameworks to better understand the limits of guarantee provisions. The objective is to not only mislead a classifier but also manipulate the certification process to generate robustness certificate guarantee for an adversarial input—*certificate spoofing*. A recent study in ICLR demonstrated crafting large perturbations can shift inputs far into regions cable of generating a certificate for an incorrect class. Our study investigates if perturbations needed to cause a misclassification and yet coax a certified model into issuing deceptive, large robustness radii can still be imperceptible. We explore the idea of region-focused adversarial examples to demonstrate imperceptible perturbations capable of spoofing certificates and achieving certification radii larger than the source class—*ghost certificates*. Extensive evaluations with *ImageNet* demonstrate the ability to effectively bypass state-of-the-art certified defenses. Our work raises new questions regarding the safe deployment of systems with certified defenses and current robustness verification methods, while underscoring the need to invest efforts to better understand and appreciate the robustness guarantees of certified models.

Certified but Fooled! Breaking Certified Defenses with Ghost Certificates

Narrative comprehension on long stories and novels has been a challenging domain attributed to their intricate plotlines and entangled, often evolving relations among characters and entities. Given the LLM's diminished reasoning over extended context and its high computational cost, retrieval-based approaches remain a pivotal role in practice. However, traditional RAG methods could fall short due to their stateless, single-step retrieval process, which often overlooks the dynamic nature of capturing interconnected relations within long-range context. In this work, we propose ComoRAG, holding the principle that narrative reasoning is not a one-shot process, but a dynamic, evolving interplay between new evidence acquisition and past knowledge consolidation, analogous to human cognition on reasoning with memory-related signals in the brain. Specifically, when encountering a reasoning impasse, ComoRAG undergoes iterative reasoning cycles while interacting with a dynamic memory workspace. In each cycle, it generates probing queries to devise new exploratory paths, then integrates the retrieved evidence of new aspects into a global memory pool, thereby supporting the emergence of a coherent context for the query resolution. Across four challenging long-context narrative benchmarks (200K+ tokens), ComoRAG outperforms strong RAG baselines with consistent relative gains up to 11% compared to the strongest baseline. Further analysis reveals that ComoRAG is particularly advantageous for complex queries requiring global comprehension, offering a principled, cognitively motivated paradigm for retrieval-based stateful reasoning.

ComoRAG: A Cognitive-Inspired Memory-Organized RAG for Stateful Long Narrative Reasoning

Most existing spatial reasoning benchmarks focus on static or globally observable environments, failing to capture the challenges of long-horizon reasoning and memory utilization under partial observability and dynamic changes. We introduce two dynamic spatial benchmarks—locally observable maze navigation and match-2 elimination—that systematically evaluate models' abilities in spatial understanding and adaptive planning when local perception, environment feedback, and global objectives are tightly coupled. Each action triggers structural changes in the environment, requiring continuous update of cognition and strategy. We further propose a subjective experience-based memory mechanism for cross-task experience transfer and validation. Experiments show that our benchmarks reveal key limitations of mainstream models in dynamic spatial reasoning and long-term memory, providing a comprehensive platform for future methodological advances.

EvoEmpirBench: Dynamic Spatial Reasoning with Agent-ExpVer

Aerial-Ground Person Re-IDentification (AGPReID) aims to extract identity-discriminative representations from heterogeneous perspectives across different platforms in complex real-world environments. However, existing methods primarily focus on visual appearance modeling and make insufficient use of semantic attribute priors, which limits their ability to bridge the aerial-ground view gap. To address this limitation, we propose a Semantic-driven Visual Progressive Refinement framework for AGPReID (SVPR-ReID), which effectively leverages textual attribute priors to guide the extraction of fine-grained visual cues. Specifically, we design a View-Decoupled Feature Extractor that incorporates view-aware textual prompts to decouple view-invariant identity features. Then, to alleviate inter-class ambiguity, we propose an Attribute-Scattered Mixture-of-Experts module that integrates attribute semantics into the visual space, thereby improving discrimination among visually similar pedestrians. Finally, we design a Context-Vision Progressive Refinement module for progressive refinement of attribute and view-invariant features, obtaining robust cross-view identity representations. In particular, we contribute a comprehensive benchmark for AGPReID, named CP2108, which contains 142,817 images of 2,108 identities annotated with 22 attributes. Notably, it includes 191 identities captured across different times, enabling both short- and long-term Re-ID evaluation, addressing the limitation of existing datasets that focus only on short-term scenarios. Extensive experimental results validate the effectiveness of our SVPR-ReID on four AGPReID datasets.

Semantic-Driven Visual Progressive Refinement for Aerial-Ground Person ReID: A Challenging Large-Scale Benchmark

Mathematical Expression Recognition (MER) has made significant progress in recognizing the simple expression, but the robust recognition of the complex mathematical expression (CMER) with numerous tokens and multiple lines remains a formidable challenge. In this paper, we first introduce CMER-Bench, a carefully constructed benchmark that categorizes expressions into three difficulty levels: normal, moderate, and complex. Leveraging CMER-Bench, we conduct a comprehensive evaluation of existing expert MER models and general-purpose multimodal large language models (MLLMs). The results reveal that while current methods perform well on normal and moderate expressions, their performance degrades significantly when handling complex mathematical expressions. In response, and considering that existing public training datasets are primarily composed of simple samples, we propose CMER-17M, a large-scale dataset specifically designed for the recognition of complex mathematical expressions. This dataset provides rich and diverse samples to support the development of accurate and robust CMER models. Furthermore, to address the challenges posed by the spatial structure of complex expressions, we introduce a novel expression representation called SML, which explicitly models the hierarchical and spatial structure of mathematical content beyond \LaTeX{} format. Based on the SML representation, we propose a specialized model named CMERNet, built upon an encoder-decoder architecture and trained on CMER-17M. Experimental results show that CMERNet, with only 0.1 billion parameters, significantly outperforms all existing expert models and MLLMs on CMER-Bench, particularly in the complex level. All data and code will be released.

Complex Mathematical Expression Recognition: Benchmark, Large-Scale Dataset and Strong Baseline

The release of open-weight large language models (LLMs) creates a tension between advancing accessible research and preventing misuse, such as malicious fine-tuning to elicit harmful content. Current safety measures struggle to preserve the general capabilities of the LLM while resisting a determined adversary with full access to the model's weights and architecture, who can use full-parameter fine-tuning to erase existing safeguards. To address this, we introduce AntiDote, a bi-level optimization procedure for training LLMs to be resistant to such tampering. AntiDote involves an auxiliary adversary hypernetwork that learns to generate malicious Low-Rank Adaptation (LoRA) weights conditioned on the defender model's internal activations. The defender LLM is then trained with an objective to nullify the effect of these adversarial weight additions, forcing it to maintain its safety alignment. We validate this approach against a diverse suite of 52 red-teaming attacks, including jailbreak prompting, latent space manipulation, and direct weight-space attacks. AntiDote is upto 27.4\% more robust against adversarial attacks compared to both tamper-resistance and unlearning baselines. Crucially, this robustness is achieved with a minimal trade-off in utility, incurring a performance degradation of upto less than 0.5\% across capability benchmarks including MMLU, HellaSwag, and GSM8K. Our work offers a practical and compute efficient methodology for building open-weight models where safety is a more integral and resilient property.

AntiDote: Bi-level Adversarial Training for Tamper-Resistant LLMs

Machine unlearning, which enables a model to forget specific data, is crucial for ensuring data privacy and model reliability. However, its effectiveness can be severely undermined in real-world scenarios where models learn unintended biases from spurious correlations within the data. This paper investigates the unique challenges of unlearning from such biased models. We identify a novel phenomenon we term "shortcut unlearning," where models exhibit an "easy to learn, yet hard to forget" tendency. Specifically, models struggle to forget easily-learned, bias-aligned samples; instead of forgetting the class attribute, they unlearn the bias attribute, which can paradoxically improve accuracy on the class intended to be forgotten. To address this, we propose CUPID, a new unlearning framework inspired by the observation that samples with different biases exhibit distinct loss landscape sharpness. Our method first partitions the forget set into causal- and bias-approximated subsets based on sample sharpness, then disentangles model parameters into causal and bias pathways, and finally performs a targeted update by routing refined causal and bias gradients to their respective pathways. Extensive experiments on biased datasets including Waterbirds, BAR, and Biased NICO++ demonstrate that our method achieves state-of-the-art forgetting performance and effectively mitigates the shortcut unlearning problem.

Content not yet available

Next from AAAI 2026

MdaIF: Robust One-Stop Multi-Degradation-Aware Image Fusion with Language-Driven Semantics

Similar lecture

Text Augmentation Using Dataset Reconstruction for Low-Resource Classification