Singapore

The rapid advancements in artificial intelligence (AI) have primarily focused on the process of learning from data to acquire knowledgeable learning systems. As these systems are increasingly deployed in critical areas, ensuring their privacy and alignment with human values is paramount. Recently, selective forgetting (also known as machine unlearning) has shown promise for privacy and data removal tasks, and has emerged as a transformative paradigm shift in the field of AI. It refers to the ability of a model to selectively erase the influence of previously seen data, which is especially important for compliance with modern data protection regulations and for aligning models with human values. Despite its promise, selective forgetting raises significant privacy concerns, especially when the data involved come from sensitive domains. While new unlearning-induced privacy attacks are continuously proposed, each is shown to outperform its predecessors using different experimental settings, which can lead to overly optimistic and potentially unfair assessments that may disproportionately favor one particular attack over the others. In this work, we present the first comprehensive benchmark for evaluating privacy vulnerabilities in selective forgetting. We extensively investigate privacy vulnerabilities of machine unlearning techniques and benchmark privacy leakage across a wide range of victim data, state-of-the-art unlearning privacy attacks, unlearning methods, and model architectures. We systematically evaluate and identify critical factors related to unlearning-induced privacy leakage. With our organized codebase and novel insights, we aim to provide a standardized tool for practitioners seeking to deploy customized unlearning applications with faithful privacy assessments.

AAAI 2026

Towards Benchmarking Privacy Vulnerabilities in Selective Forgetting with Large Language Models

selective forgetting

privacy vulnerability

machine unlearning

large language models

## Welcome to EMNLP 2022!
I am delighted to welcome you to EMNLP 2022! I believe this conference has been complicated beyond any precedent. Over the past year, it’s been thrilling to see the organization team approach each new puzzle with creativity and enthusiasm. We hope that those participating in Abu Dhabi as well as those joining remotely will leave the conference feeling newly inspired by the program and newly connected to our ever-growing community. Following EMNLP 2021 and major NLP conferences since, EMNLP 2022 is “hybrid,” serving both virtual and in-person participants.

Our key innovations for EMNLP 2022 include:

* EMNLP 2022 is “hybrid” in a second sense, as well: we allowed both direct and rolling review paper submissions, building on the pilot experiment of EMNLP 2021, which considered a small number of ARR submissions. 
* Familiar from NAACL but new to EMNLP, we’ve added an industry track.
* During the conference, “portals” will link virtual poster sessions to in-person conference participants during poster sessions each day.
* The first ACL-family conference in the United Arab Emirates.

 *Message from Noah A. Smith, University of Washington and Allen Institute for AI, Seattle, Washington, USA* 
***EMNLP 2022 General Chair***
 
[![](https://assets.underline.io/uploads/markdown_image/1/image/9eec7d4a287ee18c278b08229290aa83.png)](https://drive.google.com/file/d/1OlPv6QBeo62VVTughj2jkiLeyHd1WnUt/view)
 
[![](https://assets.underline.io/uploads/markdown_image/1/image/a3db7a768409f05192210d98601edb25.png)](https://emnlp2022.rocket.chat/)

To access this site you need to register. Please register [here](https://2022.emnlp.org/registration/).

Register here

EMNLP 2022

Welcome!
EMNLP 2022 will take place in Abu Dhabi from December 7th to December 11th, 2022. And it will be held in hybrid mode, both online and offline.

technical paper

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.

In this New Faculty Highlights, I begin with the reward free exploration that learns broad state and skill coverage with intrinsic rewards and remains robust under misspecification during efficient finetuning; guided generation methods that preserve the prior policy and mitigate reward hacking; and AI for science and healthcare, including practical RL for autonomous laboratories and automatic diagnosis. Building
on impacts evidenced by publications, adoption, and awards. My future work will pursue imitation learning and contextual multi task RL that connect behavioral cloning with interactive policies without explicit reward design; personalized and multi-tasked offline to online adaptation with in-context demonstrations. In parallel, I am broadening the impact of AI for science and healthcare through existing collaborations. I will close with a talk that surveys these results and outlines an agenda for reinforcement learning without explicit rewards.

Reinforcement Learning Without Explicit Rewards: Theory and Practice

The emergent capabilities of large language models (LLMs) have prompted interest in using them as surrogates for human subjects in opinion surveys. However, prior evaluations of LLM-based opinion simulation have relied heavily on costly, domain-specific survey data, and mixed empirical results leave their reliability in question. To enable cost-effective, early-stage evaluation, we introduce a quality control assessment designed to test the viability of LLM-simulated opinions on Likert-scale tasks without requiring large-scale human validation. This assessment comprises two key tests: logical consistency and alignment with stakeholder expectations, offering a low-cost, domain-adaptable validation tool. We apply our quality control assessment to an opinion simulation task relevant to AI-assisted content moderation and fact-checking workflows---a socially impactful use case---and evaluate seven LLMs using a baseline prompt engineering method (backstory prompting), as well as fine-tuning and in-context learning variants. None of the models or methods pass the full assessment, revealing several failure modes. We conclude with a discussion of the risk management implications and release TopicMisinfo, a benchmark dataset with paired human and LLM annotations, to support future research.

Should You Use LLMs to Simulate Opinions? Quality Checks for Early-Stage Deliberation

Palm vein recognition has emerged as a promising biometric technology, yet its development remains constrained by the scarcity of large-scale publicly available datasets. Several methods of palm vein image generation have been proposed to address this issue. These methods usually focus on the anatomical realism of palm vein patterns, but overlook the biophysical correlation between identities and vein patterns, particularly in simulating identity-specific vein contrast. To tackle this limitation, we propose a novel biophysics-driven synthesis method. 
Our method constructs a 3D palm vascular tree via established modeling method. Then, a projection model is proposed to map the 3D tree into 2D space to derive palm vein patterns. The projection model is based on skin spectral absorption and simulates the natural attenuation of light passing through the skin using a layer integration method. For different identities, we sample different skin parameters, resulting in varying degrees of attenuation. This method effectively simulates the variation in vein contrast across different identities. Furthermore, we introduce a conditional diffusion model that uses the projected patterns as identity conditions to generate palm vein images. To the best of our knowledge, this is the first palm vein generation method based on the diffusion model. 
Experimental results demonstrate that our method not only outperforms existing methods, but also enables a recognition model trained on our synthetic data to achieve superior performance compared to a model trained on real-world data at a scale of 2,000 IDs under an open-set protocol with a TAR@FAR=1:1 of 1e-4.

LSAP-PV: High-Fidelity Palm Vein Image Synthesis via Layered Spectral Absorption Projection-Guided Diffusion Model

Multi-source knowledge graph completion (MKGC) seeks to predict missing triples in a target KG by leveraging triples from multi-source KGs (e.g., different languages or domains). 
Existing studies typically learn and fuse multi-source KG representations solely with alignments or fusion modules, which can be affected by redundant information within KGs.
This issue can conceal task-relevant information in representations, impeding further improvements when scaling to numerous KGs.
To this end, we propose IMKGC, an information-theoretic MKGC framework to learn minimal sufficient representations. 
In particular, IMKGC learns entity representations explicitly preserving endogenous contextual information within each KG, exogenous complementary information from other KGs, and consistent information of equivalent entities, while suppressing redundant information through variational constraints. 
Furthermore, we achieve compressed relation representations with a devised relation reasoning decoder that captures relatedness among relations, also improving triple prediction.
Extensive experiments on 14 KGs across three multilingual and multi-domain benchmarks demonstrate that IMKGC significantly outperforms previous state-of-the-art methods, especially in redundant scenarios.
Our code will be released at \url{https://xxx} for the research community and now in the supplementary material.

Information-Theoretic Minimal Sufficient Representation for Multi-Domain Knowledge Graph Completion

World-model-based reinforcement learning achieves high sample efficiency by learning from imagined rollouts. However, its success critically depends on the accuracy of the learned world model, which is prone to producing unrealistic or hallucinated rollouts when queried beyond its domain of competence. These flawed predictions can trap the agent in a vicious cycle: by misleading exploration toward implausible or uninformative regions, they degrade the quality of collected data, which in turn corrupts policy learning with inaccurate rollouts.
To break this cycle, we introduce the notion of a knowledge boundary—the region within which the world model provides reliable predictions—and propose a unified framework that both identifies and leverages this boundary. Concretely, we approximate the boundary using model uncertainty, quantified via disagreement across an ensemble of lightweight predictors, which serves as a practical proxy. This uncertainty signal is used in two complementary ways: as an intrinsic reward to guide exploration toward under-explored yet learnable regions, and as a dynamic filter to exclude unreliable imagined rollouts from policy optimization.
Extensive experiments across diverse benchmarks—including CARLA, DeepMind Control Suite, Atari, and MemoryMaze—demonstrate that our approach consistently outperforms prior state-of-the-art methods.

Perceiving the Knowledge Boundary: Uncertainty-Guided Exploration and Imagination for World Models

Novel object synthesis by integrating distinct textual concepts from diverse categories remains a significant challenge in Text-to-Image (T2I) generation. Existing methods often suffer from insufficient concept mixing, lack of rigorous evaluation, and suboptimal outputs—manifesting as conceptual imbalance, superficial combinations, or mere juxtapositions. To address these limitations, we propose \textbf{Reinforcement Mixing Learning (RMLer)}, a framework that formulates cross-category concept fusion as a reinforcement learning problem: mixed features serve as states, mixing strategies as actions, and visual outcomes as rewards. Specifically, we design an MLP-policy network to predict dynamic coefficients for blending cross-category text embeddings. We further introduce visual rewards based on (1) semantic similarity and (2) compositional balance between the fused object and its constituent concepts, optimizing the policy via proximal policy optimization. At inference, a selection strategy leverages these rewards to curate the highest-quality fused objects. Extensive experiments demonstrate RMLer’s superiority in synthesizing coherent, high-fidelity objects from diverse categories, outperforming existing methods. Our work provides a robust framework for generating novel visual concepts, with promising applications in film, gaming, and design.

RMLer: Synthesizing Novel Objects Across Diverse Categories via Reinforcement Mixing Learning

Sketch person re-identification aims to match hand-drawn sketches with RGB surveillance images, yet remains challenging due to severe modality discrepancies and limited labeled data. To address this, we propose KTCAA, a theoretically grounded and interpretable framework for few-shot cross-modal transfer learning. From the perspective of generalization bounds, we identify two key controllable factors essential to minimizing target domain error: (1) domain discrepancy, which reflects the difficulty of aligning source and target distributions in the feature space; and (2) perturbation invariance, which measures the model’s robustness to cross-modal variations. To address these challenges, we design two corresponding modules: (1) Alignment Augmentation (AA), which applies localized sketch-style transformations to simulate target modality distributions, introducing slight but meaningful feature shifts that guide gradual distribution alignment during training; and (2) Knowledge Transfer Catalyst (KTC), which enhances perturbation invariance by generating worst-case adversarial modality perturbations and enforcing output consistency under such perturbations. These modules are jointly optimized within a meta-learning framework that transfers alignment knowledge from RGB-rich domains to sketch scenarios. Experiments on multiple benchmarks demonstrate that KTCAA achieves state-of-the-art performance, particularly under limited data and cross-domain transfer settings.

A Theory-Inspired Framework for Few-Shot Cross-Modal Sketch Person Re-Identification

The Aperiodic Tiling Complements Problem (ATCP) involves finding the full set of possible aperiodic complements of a given pattern.
This has become a classic problem in music theory, with some recent attempts to model it using Integer Linear Programming (ILP) and Boolean Satisfiability (SAT) frameworks.
In this paper, we propose and compare different models of ATCP encoded in Constraint Programming (CP).
The most effective approach admits two phases: a first one that allows us to merge (join) several subsets of linear constraints under the form of tables with large arity, and a second one that advantageously exploits the generated tables to discard periodic tiling complements.
Our experimental results show that our approach significantly outperforms the state-of-the-art, solving every instance of a classical benchmark (standard Vuza rhythms for canons
with period set to 420 and 900) in a time within 5 seconds to 2 minutes (except the largest instance being solved in 18 minutes).

Aperiodic Tiling and Rhythmic Canons: A CP Journey

Recommendation systems commonly face selection bias from missing-not-at-random (MNAR) data. While propensity-based methods such as inverse propensity scoring (IPS) and doubly robust (DR) estimators are widely used, their existing variant methods typically optimize only one or two evaluation criteria, neglecting their interdependencies and inherent trade-offs, limiting robustness.
To address this, we propose a bin-based \textbf{Unified Propensity Optimization (UPO)} framework that integrates and upper-bounds four widely-used criteria by revealing a fundamental structural commonality among their corresponding metrics, Expected Calibration Error (calibration), Bias Minimizing Squared Error (feature balancing), DR bias, and DR variance, showing that each metric can be represented as a feature- and model-driven adaptive weights multiplied by propensity estimation errors. Based on this unified structure, we design a minimax optimization algorithm employing adversarial weighting well addressing the trade-off between all four criteria . Theoretically, we demonstrate that minimizing the UPO loss effectively controls the original global-scale metrics, ensuring their simultaneous improvements without incurring additional bias, and achieving reduced variance compared to naïvely adding up multiple penalization terms.
Empirically, experiments on semi-synthetic and real-world datasets validate UPO’s efficacy, demonstrating superior performance compared to state-of-the-art methods with minor computational overhead.

Unified Minimax Optimization Framework for Propensity Score Estimation in Debiased Recommendation

Modern task-oriented dialogue (TOD) systems increasingly rely on large language model (LLM) agents, leveraging Retrieval-Augmented Generation (RAG) and long-context capabilities for long-term memory utilization. However, these methods are primarily based on semantic similarity, overlooking task intent and reducing task coherence in multi-session dialogues. To address this challenge, we introduce **MemGuide**, a two-stage framework for intent-driven memory selection. (1) **Intent‑Aligned Retrieval** matches the current dialogue context with stored intent descriptions in the memory bank, retrieving QA‑formatted memory units that share the same goal. (2) **Missing‑Slot Guided Filtering** employs a chain‑of‑thought slot reasoner to enumerate unfilled slots, then uses a fine‑tuned LLaMA‑8B filter to re‑rank the retrieved units by marginal slot‑completion gain. The resulting memory units inform a proactive strategy that minimizes conversational turns by directly addressing information gaps. Based on this framework, we introduce the **MS-TOD**, the first multi-session TOD benchmark comprising 132 diverse personas, 956 task goals, and annotated intent-aligned memory targets, supporting efficient multi‑session task completion. Evaluations on MS-TOD show that MemGuide raises the task success rate by 11\% (88\%→99\%) and reduces dialogue length by 2.84 turns in multi-session settings, and maintains parity with single‑session benchmarks.

Downloads

Next from AAAI 2026

Reinforcement Learning Without Explicit Rewards: Theory and Practice

Similar lecture

Knowledge Unlearning for Mitigating Privacy Risks in Language Models