Singapore

Automated planning is a long-standing challenge in AI research, which requires domain models that define the state variables and actions in terms of preconditions and effects. However, the handcrafting of such domain models is widely recognized to be a time-consuming and error-prone process. Therefore, theories and algorithms for automatically learning domain models for planning have been developed over the years. The goal of this tutorial is to provide participants with an overview of state-of-the-art approaches for learning planning domain models under different assumptions, and to enable participants to utilize publicly available frameworks and tools for domain model learning.

The tutorial first introduces the basic concepts for planning domain learning, and situates the problem within the broader context of AI research, relating it to other fields such as Reinforcement Learning (RL) and Knowledge Representation (KR). Then, the tutorial provides an overview of state-of-the-art approaches for learning domain models assuming a given symbolic state representation. The focus is on approaches that learn action preconditions and effects from an input set of trajectories generated by executing actions and observing their effects. The discussed approaches are compared w.r.t. their pros and cons, and their theoretical properties. Subsequently, the tutorial discusses domain model learning approaches that do not assume a symbolic state representation. Such approaches additionally learn to encode state subsymbolic observations (e.g., RGB images) into symbolic representations suitable for planning. A hands-on session guides the participants through the usage of open-source frameworks and tools for domain model learning and evaluation. The tools considered are useful and readily accessible resources for practitioners and researchers who aim to apply domain model learning techniques in possibly different research areas. The tutorial ends by describing methods for online domain model learning, and integrations with RL agents and LLM-based agents.

Participants are expected to have only a basic understanding of automated planning and symbolic domain representation, i.e., that a symbolic state is defined by a set of state variables, and a symbolic action is specified by a set of preconditions and effects. Further information about the tutorial is available at https://domain learning.github.io.

AAAI 2026

Domain Model Learning for Automated Planning: Offline Learning of Action 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.

tutorial

Post-training alignment is crucial for refining the reasoning capabilities of Large Language Models (LLMs). A dominant paradigm for this involves optimizing the model's policy using reinforcement learning, powered by techniques such as Proximal Policy Optimization (PPO), Direct Preference Optimization (DPO), and Group Relative Policy Optimization (GRPO). The success of these methods, whether using an explicit reward model or optimizing directly on preference data, is critically dependent on the quality of the guiding signal. However, these signals are conventionally derived from task-specific outcomes, such as correctness in math or fluency in summarization. This approach often limits the model's ability to generalize its reasoning skills across diverse domains and can lead to reward hacking or model collapse. This paper challenges this outcome-based paradigm by introducing a conceptual framework, GRIT (Generalizable Reasoning via Intrinsic Traits). This novel framework aims to shift the emphasis from rewarding what the model answers to how it reasons. To accomplish this, we define a set of universal, task-agnostic traits of sound cognition inspired by human reasoning. These intrinsic traits are encoded as distinct reward components: (1) ensuring sequential logical coherence, (2) penalizing cyclic or redundant reasoning, (3) rewarding successful and integrated tool utilization, and (4) maintaining semantic alignment with the user's query. By fine-tuning an LLM to optimize for these intrinsic traits, we hypothesize that the model will develop a more robust and generalizable cognitive process.

Rethinking Reward Models! A Conceptual Framework for Enhancing LLM Reasoning through Intrinsic Traits

Online continual learning (OCL) methods adapt to changing environments without forgetting past knowledge. Similarly, online time series forecasting (OTSF) is a real-world problem where data evolve in time and success depends on both rapid adaptation and long-term memory. Indeed, time-varying and regime-switching forecasting models have been extensively studied, offering a strong justification for the use of OCL in these settings. Building on recent work that applies OCL to OTSF, this paper aims to strengthen the theoretical and practical connections between time series methods and OCL. First, we reframe neural network optimization as a parameter filtering problem, showing that natural gradient descent is a score-driven method and proving its information-theoretic optimality. Then, we show that using a Student’s t likelihood in addition to natural gradient induces a bounded update, which improves robustness to outliers. Finally, we introduce Natural Score-driven Replay (NatSR), which combines our robust optimizer with a replay buffer and a dynamic scale heuristic that improves fast adaptation at regime drifts. Empirical results demonstrate that NatSR achieves stronger forecasting performance than more complex state-of-the-art methods.

Online Continual Learning for Time Series: a Natural Score-driven Approach

Current Large Language Model (LLM) training paradigms, while effective at pattern matching and knowledge retrieval, often fall short of replicating the complex, adaptive and generalizable reasoning characteristic of human intelligence. We argue that this arises from a fundamental disconnect between the static, data-driven training of LLMs and the dynamic, lifelong learning process inherent to human cognitive development. To bridge this gap, we introduce Learn-Master-Teach Tuning (LMT) , a novel, end-to-end training framework that simulates the complete human `student-to-teacher' life-cycle. Our paradigm guides the model through a comprehensive developmental trajectory, from a beginner who is trying to adjust and learn a curriculum, to a seasoned educator capable of lifelong learning and knowledge synthesis. By situating learning within a comprehensive, cognitive-inspired framework, we explore two fundamental research questions: Can the deep simulation of a human persona, in this case, a developing academic, lead to a genuine replication of that persona's capabilities? And, does this student-teacher life-cycle offer a superior training paradigm for fostering robust and generalizable reasoning in LLMs? We present the complete LMT methodology and position it within the landscape of existing training paradigms, arguing that by emulating the human journey of learning, we can unlock new frontiers in artificial general intelligence such as lifelong learning and continual adaptation with history retention. Initial results using small language models in the challenging domain of differential equations already demonstrate significant improvements of about 33% over baselines indicating the transformative potential of the proposed human-pedagogy inspired framework.

Human-Pedagogy Inspired LLM Fine-Tuning Paradigm for Lifelong Leaning and Continual Adaptation

Given the recent tremendous success of large language models (LLMs), there has been an increasing trend in applying them to several down-stream tasks via fine-tuning (FT). However, there persist two significant challenges in FT - 1) curation of high-quality task-specific data and 2) expensive time-consuming model adaptation via gradient descent optimization. To mitigate these limitations, we leverage prior works in large-scale parameter generation for LLMs and propose Self-Optimizing Lifelong Autonomous Reasoner (SOLAR) to open up a new paradigm of parameter-level meta learning, thereby serving a critical advance in the AI Scientist domain. SOLAR is an open-ended autonomous foundation-model-based agent which is capable of self-improvement by discovering and learning the rich meta-knowledge information present in large neural network weights, thereby enabling efficient adaptation of LLMs by parameter-level weight modifications to unseen domains as well. To achieve this, SOLAR has a multi-level reinforcement learning approach to train models for efficient understanding of parameter space and offers higher degree of flexibility compared to prior works on self-evolution by providing freedom to the model for choosing its own adaptation strategy thereby breaking the scaling solely through data regime. Early experiments demonstrate the superior performance of SOLAR in the common-sense reasoning domain as it outperforms task-specific FT by 23.6% on average and even some of the most recent works in parameter generation (by 10.4%), model merging (by 24.3%) and test-time learning (by 25.2%) as well out-of-domain tasks such as coding, social, logical and mathematical reasoning as well.

SOLAR : A Self-Optimizing Open-Ended Autonomous Agent for Lifelong Learning and Continual Adaptation

Domain Model Learning for Automated Planning: Hands-on Session

Domain Model Learning for Automated Planning: Active Learning and Open Challenges 

LH03: The Verification of Neural Networks Competition (VNN-COMP): A Lab for Benchmark Proposers, Verification Tool Participants, and the Broader AI Community

Associative Memories (AMs) like Hopfield Networks are elegant models for describing fully recurrent neural networks that store and retrieve information. Recent theoretical advances have revealed deep connections between AMs and modern AI architectures such as Transformers and diffusion models, opening new possibilities for interpreting and designing neural networks. This tutorial provides an approachable introduction to AMs through three components: theoretical foundations from classical to Dense Associative Memories, hands-on implementations using a universal Lagrangian framework for designing AMs (including coding Transformers as AMs), and practical applications connecting AMs to kernel methods, density estimation, and deep clustering.

We provide a technical, hands-on level tutorial for the general attendee of AAAI i.e., graduate students, industry researchers, faculty, and developers. However, we expect our tutorial to additionally appeal to physicists, neuroscientists, and other researchers interested in brain-inspired models of computation. From a student’s perspective, the ordered priorities of this tutorial are to: (1) build intuition, (2) understand the fundamental tooling needed to work with these systems, and (3) engage with simple proofs that highlight their elegant behavior.

The tutorial is accompanied by a website containing links to all lecture notes, code examples, and presentation slides.

Modern Methods in Associative Memory

As generative AI systems such as large language models (LLMs), diffusion models, and vision–language models become widely deployed, the ability to “unlearn” sensitive or harmful information has become essential for AI safety, compliance, and trust. Machine unlearning (MU) enables models to selectively forget undesirable knowledge, ranging from private data and copyrighted content to hazardous instructions, but current methods remain fragile and easily reversed. This tutorial offers a structured exploration of the science and practice of machine unlearning, highlighting both recent advances and persistent challenges. We begin by motivating the need for MU through real-world safety applications, then clarify what distinguishes faithful forgetting from shallow suppression. We review major approaches, such as data-driven, model-based, and input-driven, and examine their tradeoffs in effectiveness, efficiency, and robustness. A central focus will be the optimization foundations of unlearning, where we present principled formulations that balance forgetting with utility preservation and enhance its robustness. We also uncover critical vulnerabilities, showing how jailbreak prompts, low-shot fine-tuning, or quantization can reactivate forgotten knowledge, and connect these failure modes to broader themes in adversarial machine learning. In addition to the foundations of MU, participants will gain practical insights through exposure to public unlearning benchmarks and toolkits. Walkthroughs of representative MU pipelines will illustrate common failure modes, stress-testing strategies, and evaluation protocols. The tutorial concludes with open research directions, including interpretability, verification, compliance-aware deployment, and unlearning for reasoning-enabled AI models and agents. Designed for a broad audience, the tutorial requires no prior familiarity with unlearning. All key concepts will be introduced with explanations, intuitive illustrations, and references. By bridging optimization, adversarial robustness, and AI safety, this tutorial charts a principled roadmap toward AI systems that can forget efficiently, robustly, and responsibly.

When AI "Forgets" for Good: The Science and Practice of Machine Unlearning for AI Safety — Progress, Pitfalls, and Prospects

The last few years have seen the rapid development of machine learning methods for natural language processing, computer vision, and scientific discovery. Many recently developed tools and cutting-edge methodologies are from the theory of optimal transport (OT) and have been particularly successful for various applications, e.g., the model architectures and learning algorithms driven by theoretical and computational optimal transport techniques. In this tutorial, we plan to introduce the AAAI community to recent advances in theoretical and computational optimal transport techniques and their applications in designing new model architectures and learning algorithms, especially those proven to be very effective in challenging tasks but are not yet well-known.

The content of the tutorial relates to several areas within the AAAI community, such as statistical machine learning, optimization algorithms, neural network design and analysis, and various cutting-edge machine learning applications, including multi-modal representation learning, structured data (like graphs and point clouds) modeling, and high-dimensional data generation. The target audience of this tutorial includes but is not limited to computer scientists, mathematicians, machine learning researchers, and multidisciplinary practitioners, from both theoretical and applied communities, and at both junior and senior levels. The participants will learn about theoretical fundamentals, practical implementations, and innovative applications of optimal transport-driven machine learning techniques, the most recent advances of related topics, as well as open problems and directions for future work.

Downloads

Next from AAAI 2026

Rethinking Reward Models! A Conceptual Framework for Enhancing LLM Reasoning through Intrinsic Traits

Similar lecture

Domain Model Learning for Automated Planning: Active Learning and Open Challenges

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

Rethinking Reward Models! A Conceptual Framework for Enhancing LLM Reasoning through Intrinsic Traits

Similar lecture

Domain Model Learning for Automated Planning: Active Learning and Open Challenges

Stay up to date with the latest Underline news!

PRESENTATIONS

CONFERENCES

COMPANY

RESOURCES

Downloads