AAAI 2026

3,427 papers

Top Research Directions

Research directions most represented by AAAI 2026 papers.

Hyperparameter Sensitivity in Complex Machine Learning Frameworks 11
Evaluating Generative AI with Large Language Models 10
Overcoming the Limits of Multi Step Language Model Reasoning 9
Preserving Model Utility During Machine Unlearning 7
Dynamic Preference Modeling in Interactive Systems 7
Factual Inaccuracy and Hallucination in Large Language Models 6
Mitigating Non-Stationary Distribution Shift in Time-Series Forecasting 6
The Empirical Validation Gap 6
3D Gaussian Splatting for Photorealistic Scene Representation 5
Autonomous Code Refinement via Execution Feedback Loops 5
Systematic Biases in Large Language Model Evaluators 5
Intermediate Reasoning Chains for Deliberative Inference 5
Mitigating the Alignment Tax in Large Language Models 5
Mitigating the Inference Latency of Chain of Thought 5
Shifting from Static Benchmarks to Dynamic Evaluation 5

Top Authors

AuthorPapers
David Scott Lewis4
Xinyan Liang4
Adam Dziedzic3
Changshuo Wang3
Jiexi Yan3
Shuaibo Li3
Yazhou Yao3
Alexey Naumov2
Amit Ranjan Trivedi2
Bo Li2
Boxun Xu2
Chaoran Feng2
Chenyu Zhang2
Eric Granger2
Fangyuan Mao2
Gaole Dai2
Guansong Pang2
Guibo Zhu2
Hanqing Wang2
Hao Yu2

Impactful Papers

Papers from AAAI 2026 that shaped research clusters.

Foundational

Papers

Title
Beyond Local Patterns: Multiscale Inconsistency Learning for Graph Anomaly Detection
Beyond Missing Data Imputation: Information-Theoretic Coupling of Missingness and Class Imbalance for Optimal Irregular Time Series Classification
Beyond Monotonicity: Revisiting Factorization Principles in Multi-Agent Q-Learning
Beyond Neuron-Level Sparsity: Achieving Faithful and Interpretable LLMs with Mixture of Decoders
Beyond Next Token Probabilities: Learnable, Fast Detection of Hallucinations and Data Contamination on LLM Output Distributions
Beyond N-grams: A Hierarchical Reward Learning Framework for Clinically-Aware Medical Report Generation
Beyond Observations: Reconstruction Error-Guided Irregularly Sampled Time Series Representation Learning
Beyond Passive Critical Thinking: Fostering Proactive Questioning to Enhance Human-AI Collaboration
Beyond Patches: Mining Interpretable Part-Prototypes for Explainable AI
Beyond Perplexity: Let the Reader Select Retrieval Summaries via Spectrum Projection Score
Beyond Plain Demos: A Demo-Centric Anchoring Paradigm for In-Context Learning in Alzheimer's Disease Detection
Beyond Quadratic: Linear-Time Change Detection with RWKV
Beyond ReAct: A Planner-Centric Framework for Complex Tool-Augmented LLM Reasoning
Beyond Semantic Features: Pixel-level Mapping for Generalized AI-Generated Image Detection
Beyond Sharpness: A Flatness Decomposition Framework for Efficient Continual Learning
Beyond Simple Edits: X-Planner for Complex Instruction-Based Image Editing
Beyond Single-Speed Reasoning: Coordinating Fast and Slow Dynamics for Efficient World Modeling
Beyond Single-Step Updates: Reinforcement Learning of Heuristics with Limited-Horizon Search
BeyondSparse: Facilitating Mamba to Enhance Cross-Domain 3D Semantic Segmentation in Adverse Weather
Beyond Static: Related Questions Retrieval Through Conversations in Community Question Answering
Beyond Step Pruning: Information Theory Based Step-level Optimization for Self-Refining Large Language Models
Beyond Superficial Forgetting: Thorough Unlearning Through Knowledge Density Estimation and Block Re-Insertion
Beyond the Black Box: Demystifying Multi-Turn LLM Reasoning with VISTA
Beyond the Horizon: Decoupling Multi-View UAV Action Recognition via Partial Order Transfer
Beyond the Lower Bound: Bridging Regret Minimization and Best Arm Identification in Lexicographic Bandits
Beyond Tokens: Dynamic Latent Reasoning via Semantic Residual Refinement
Beyond Training-time Poisoning: Component-level and Post-training Backdoors in Deep Reinforcement Learning
Beyond Transcription: Mechanistic Interpretability in ASR
Beyond Translation: A Decomposed Collaborative Reasoning Framework Harnessing LLMs and Symbolic Solvers
Beyond Wide-Angle Images: Structure-to-Detail Video Portrait Correction via Unsupervised Spatiotemporal Adaptation
BhashaKritika: Building Synthetic Pretraining Data at Scale for Indic Languages
Bias Association Discovery Framework for Open-Ended LLM Generations
Bias-Restrained Prefix Representation Finetuning for Mathematical Reasoning
BiCA: Effective Biomedical Dense Retrieval with Citation-Aware Hard Negatives
BiCycle: Group-wise Recursive Transformer Based on ASR Mechanism
Bid Farewell to Seesaw: Towards Accurate Long-Tail Session-Based Recommendation via Dual Constraints of Hybrid Intents
Bidirectional Bounded-Suboptimal Heuristic Search with Consistent Heuristics
Bidirectional Channel-selective Semantic Interaction for Semi-Supervised Medical Segmentation
Bidirectional Counterfactual Distillation for Review-Based Recommendation
Bidirectional Noise Injection: Enhancing Diffusion Models via Coordinated Input-Output Perturbation
BiHiTo: Biomolecular Hierarchy-inspired Tokenization
Bilevel MCTS for Amortized O(1) Node Selection in Classical Planning
Bi-level Personalization for Federated Foundation Models: A Task-vector Aggregation Approach
Binary-Gaussian: Compact and Progressive Representation for 3D Gaussian Segmentation
Binary Message Passing for Generalizable Semi-Supervised Graph Anomaly Detection
BioDPP: Dynamic Prompt Policy Learning for Biomedical Vision-Language Models
BiO-HMC: Dynamic Human-Machine Collaboration for Consensus Decision-Making via Bilevel Optimization
Biologically-Inspired Evolutionary Domain Symbiosis for Few-shot and Zero-shot Point Cloud Semantic Segmentation
BIQ: Bisection Interval Quantization for Communication-efficient Federated Learning
Bi-Spectrum Distillation: Addressing Spectral Mismatch in ANN-SNN Knowledge Transfer

« prev Page 7/69 next »