Singapore

My research lies at the intersection of causality,
reinforcement learning, world models, and multi-agent
systems. I aim to develop causal foundation world models
that enable agents to interpret the past, reason about the
future, and act reliably in dynamic, non-stationary, and
open-ended environments. My work spans causal
representation learning (e.g., CausalVAE), causal reasoning
in large language models, and causality-driven exploration
in open-ended worlds. These contributions have appeared in
leading venues such as NeurIPS, ICML, ICLR, CVPR, and KDD,
and have been recognized through over 770 citations and the
Rising Star in AI award (2024). Looking forward, my agenda
focuses on scalable, trustworthy causal world models for
healthcare, robotics, scientific discovery, and digital
systems.

AAAI 2026

Toward Causal Foundation World Models: From Representation to Decision-Making

agent decision making

foundation model

causality

world models

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.

technical paper

Test-time prompt tuning for vision-language models has demonstrated impressive generalization capabilities under zero-shot settings. However, tuning the learnable prompts solely based on unlabeled test data may induce prompt optimization bias, ultimately leading to suboptimal performance on downstream tasks. In this work, we analyze the underlying causes of prompt optimization bias from both the model and data perspectives. In terms of the model, the entropy minimization objective typically focuses on reducing the entropy of model predictions while overlooking their correctness. This can result in overconfident yet incorrect outputs, thereby compromising the quality of prompt optimization. On the data side, prompts affected by optimization bias can introduce misalignment between visual and textual modalities, which further aggravates the prompt optimization bias. To this end, we propose a Doubly Debiased Test-Time Prompt Tuning method. Specifically, we first introduce a dynamic retrieval-augmented modulation module that retrieves high-confidence knowledge from a dynamic knowledge base using the test image feature as a query, and uses the retrieved knowledge to modulate the predictions. Guided by the refined predictions, we further develop a reliability-aware prompt optimization module that incorporates a confidence-based weighted ensemble and cross-modal consistency distillation to impose regularization constraints during prompt tuning. Extensive experiments across 15 benchmark datasets involving both natural distribution shifts and cross-datasets generalization demonstrate that our method consistently outperforms baselines, validating its effectiveness in mitigating prompt optimization bias.

Doubly Debiased Test-Time Prompt Tuning for Vision-Language Models

Graph anomaly detection is emerging as a critical technology for addressing increasingly complex and dynamic risk environments. 
Although unsupervised graph anomaly detection has advanced under the graph representation learning, directly applying these paradigms remains fundamentally misaligned with anomaly detection objectives.
In this work, we highlight two key insights: graph neural networks are often suboptimal as feature extractors due to neighborhood aggregation diluting anomaly signals, and reliance on local inconsistency mining is inadequate for comprehensive anomaly detection, as it often fails to identify anomalies hidden within camouflaged communities.
Based on these insights, we propose multiscale inconsistency learning for graph anomaly detection (MI-GAD), a novel framework that integrates both local and global anomaly signals. Specifically, individual node representations are projected onto a common hypersphere to ensure uniformity. At the local scale, the graph structure is leveraged for affinity-aware modeling via group discrimination. At the global scale, we introduce node deviation, a metric that distinguishes anomalies by optimizing representation centers. This unified approach enables robust and comprehensive detection of diverse graph anomalies.
Experiments on seven real datasets demonstrate that our method consistently outperforms state-of-the-art baselines in both effectiveness and scalability.

Beyond Local Patterns: Multiscale Inconsistency Learning for Graph Anomaly Detection

In multi-view multi-label (MVML) classification, each sample is represented by multiple heterogeneous views and annotated with multiple labels. Existing methods typically exploit pairwise semantic relationships to mine intra-view correlations and align inter-view features for generating structural representations. However, these methods ignore the direct expression of high-order semantic similarities and alignments from a group perspective, which necessitates multi-step aggregation for subsequent feature fusion, leading to the inefficient and incomplete integration of key semantic information. To overcome this limitation, we propose a novel hypergraph-based MVML method with Adaptive High-Order Semantic Fusion (HyperAHSF), which leverages hypergraphs to adaptively model group-level semantic similarities within each view and group-level semantic alignments across different views, enabling more effective feature fusion. Specifically, we first construct view-specific hyperedges by selecting multiple groups of node representations exhibiting high semantic similarity, which captures the group-level semantic similarities within each view, forming view-specific hypergraphs. Furthermore, we establish cross-view hyperedges to connect the multi-view node representations of each sample, which characterizes the group-level semantic alignments across different views, accordingly forming a unified multi-view hypergraph. Afterwards, we employ hypergraph neural networks to efficiently aggregate view-specific information and consensus information from their corresponding hypergraphs via group-level message passing. During the passing process, we impose a label-driven contrastive loss on the consensus information to encourage these representations to cluster toward their corresponding class prototypes, enhancing their discriminability. Finally, the consensus information together with the view-specific information is jointly integrated for multi-label classification. Extensive experiments demonstrate that HyperAHSF outperforms other state-of-the-art methods.

Hypergraph-Based Multi-View Multi-Label Classification via Adaptive High-Order Semantic Fusion

Controllable diffusion models have shown great promise in visual generation by conditioning on diverse structural inputs, such as edges and depth. However, existing methods either require separate models for each control signal or rely on unified architectures with entangled representations, resulting in poor generalization and high computational cost when adapting to new conditions.
We introduce \textbf{DivControl}, a decomposable pre-training framework for unified and efficient conditional generation. By applying knowledge diversion to the condition encoder, DivControl explicitly disentangles condition-agnostic semantics (learngenes) from condition-specific modulations (tailors) via structured SVD-based decomposition. A dynamic soft routing mechanism selects relevant tailors based on semantic task embeddings, enabling zero-shot generalization and parameter-efficient adaptation to novel control types. To further enhance feature alignment, we incorporate a representation alignment loss that anchors early condition features to high-level semantic embeddings.
Extensive experiments on a 20-condition benchmark demonstrate that DivControl achieves state-of-the-art controllability with \textbf{300$\times$ less training cost} than prior methods. It improves FID by \textbf{2.33} on seen conditions and shows strong zero-shot and few-shot performance on unseen modalities, highlighting its scalability, modularity, and transferability in controllable diffusion.

DivControl: Knowledge Diversion for Controllable Image Generation

Learning decision policies from confounded observational data is a challenging task in causal inference, as unobserved confounders can lead to biased or suboptimal actions when relying solely on machine learning models. A synergistic approach is learning to defer, which decides when to act itself and when to defer to a human expert with access to unobserved information. However, constructing the learning target, which defines the probability of choosing each action or deferral, remains a core challenge. To address this, we propose causal-target-based learning to defer (CTLD) framework, where the causal target is constructed from sharp bounds on potential outcomes. Specifically, the degree of overlap between these bounds determines the probability of deferral, while their relative positions and widths define the probabilities over actions. CTLD aligns model predictions with this causal target to make probabilistic decisions over actions and deferral. We present comprehensive theoretical guarantees for the learned policy and demonstrate the effectiveness of CTLD on synthetic and semi-synthetic datasets.

A Causal Target for Learning to Defer Under Hidden Confounding

Vision-Language Models (VLMs) have been widely used in various visual recognition tasks due to their remarkable generalization capabilities. 
As these models grow in size and complexity, fine-tuning becomes costly, emphasizing the need to reuse adaptation knowledge from 'weaker' models to efficiently enhance 'stronger' ones.
However, existing adaptation transfer methods exhibit limited transferability across models due to their model-specific design and high computational demands.
To tackle this, we propose Transferable Model-agnostic adapter (TransMiter), a light-weight adapter that improves vision-language models 'without backpropagation'.
TransMiter captures the knowledge gap between pre-trained and fine-tuned VLMs, in an 'unsupervised' manner.
Once trained, this knowledge can be seamlessly transferred across different models without the need for backpropagation.
Moreover, TransMiter consists of only a few layers, inducing a negligible additional inference cost.
Notably, supplementing the process with a few labeled data further yields additional performance gain, often surpassing a fine-tuned stronger model, with a marginal training cost.
Experimental results and analyses demonstrate that TransMiter effectively and efficiently transfers adaptation knowledge while preserving generalization abilities across VLMs of different sizes and architectures in visual recognition tasks.

Transferable Model-agnostic Vision-Language Model Adaptation for Efficient Weak-to-Strong Generalization

Recent advances in Multimodal Large Language Models (MLLMs) have enabled joint reasoning over financial textual and visual inputs. However, they still struggle with financial terminology, logical consistency, and numeric computations. Moreover, while closed-source models perform well on reasoning tasks, their high inference costs limit the scalable usage in real-world financial applications. We thus propose a cost-effective framework, CLER, combining contrastive retrieval with step-wise reflection, to improve reasoning performance. Also, the reasoning cost only generated in test stage when using commercial large models. CLER leverages **FinErrorSet**, a dataset of 8,000+ mistake-correction pairs from diverse open-source MLLMs. A fine-grained retriever is trained to identify structurally relevant errors for self-correction through individual reflection. Experiments on three multimodal benchmarks show that CLER consistently outperforms other baselines. To our knowledge, CLER is the first framework to use cross-model errors for financial reasoning.

CLER: Improving Multimodal Financial Reasoning by Cross-MLLM Error Reflection

We study a sequential decision-making model where a set of items is repeatedly matched to the same set of agents over multiple rounds. The objective is to determine a sequence of matchings that either maximizes the utility of the least advantaged agent at the end of all rounds (optimal) or at the end of every individual round (anytime optimal). We investigate the computational challenges associated with finding (anytime) optimal outcomes and demonstrate that these problems are generally computationally intractable. However, we provide approximation algorithms, fixed-parameter tractable algorithms, and identify several special cases whereby the problem(s) can be solved efficiently. Along the way, we also establish characterizations of Pareto-optimal/maximum matchings, which may be of independent interest to works in matching theory and house allocation.

Fairness in Repeated Matching: A Maximin Perspective

Denoising higher-resolution latents using a pre-trained U-Net often results in repetitive and disordered image patterns. In this work, we are motivated to reveal the intrinsic cause of such pattern disruption in high-resolution image generation. 
Through theoretical analysis and empirical studies, we reveal that the pre-trained U-Net fails to provide sufficient positional information for tokens at high-resolution. Specifically, 1) zero-padding serves as a critical mechanism for position encoding but lacks robustness across varying resolutions; and 2) tokens located farther from the feature map boundaries have increasing difficulty acquiring positional awareness, leading to pattern disruptions. 
Inspired by these findings, we propose a novel training-free approach for high-resolution generation, introducing a Progressive Boundary Complement (PBC) method. It creates dynamic virtual image boundaries inside the feature map to supplement position information at high resolution, enabling high-quality and rich-content high-resolution image synthesis. Extensive experiments show that our method significantly improves high-resolution image synthesis in terms of visual quality and content richness, achieving state-of-the-art performance.

Exploring Position Encoding Mechanism in Diffusion U-Net for Training-free High-resolution Image Generation

Definite descriptions are expressions of the form „the unique x satisfying property C,” which allow reference to objects through their distinguishing characteristics. They play a crucial role in ontology and query languages, offering an alternative to proper names (IDs), which lack semantic content and serve merely as placeholders. 

In this paper, we introduce two extensions of the well-known description logic ALC with local and global definite descriptions, denoted ALCiL and ALCiG, respectively. We define appropriate bisimulation notions for these logics, enabling an analysis of their expressiveness. We show that although both logics share the same tight ExpTime complexity bounds for concept and ontology satisfiability, ALCiG is strictly more expressive than ALCiL. Moreover, we present tableau-based decision procedures for satisfiability in both logics, provide their implementation, and report on a series of experiments. The empirical results demonstrate the practical utility of the implementation and reveal interesting correlations between performance and structural properties of the input formulas.

Downloads

Next from AAAI 2026

Doubly Debiased Test-Time Prompt Tuning for Vision-Language Models

Similar lecture

Eliciting Causal Knowledge from Agents

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

Doubly Debiased Test-Time Prompt Tuning for Vision-Language Models

Similar lecture

Eliciting Causal Knowledge from Agents

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES

Downloads