Ax Dip Roy 3/20/2026

Causal Intervention Framework for Variational Auto Encoder Mechanistic Interpretability

Causal intervention framework for interpreting Variational Autoencoders mechanistically, addressing interpretability of generative models.

Ax Xiang Shi, Rui Zhang, Jiawei Liu, Yinpeng Liu, Qikai Cheng, Wei Lu 3/20/2026

Modality Equilibrium Matters: Minor-Modality-Aware Adaptive Alternating for Cross-Modal Memory Enhancement

Shapley Value-based alternating training framework for multimodal fusion that balances dominant and minor modalities.

Ax Antonio Ferrara, Francesco Cozzi, Alan Perotti, Andr\'e Panisson, Francesco Bonchi 3/20/2026

Size-adaptive Hypothesis Testing for Fairness

Statistical framework for fairness testing in algorithmic systems that accounts for sampling error and handles intersectional demographic analysis.

Ax Tongtian Zhu, Tianyu Zhang, Mingze Wang, Zhanpeng Zhou, Can Wang 3/20/2026

On the Surprising Effectiveness of a Single Global Merging in Decentralized Learning

Analysis of communication scheduling in decentralized learning showing benefits of concentrating synchronization in later training stages.

Ax Kyeongjin Ahn, Sungwon Han, Seungeon Lee, Donghyun Ahn, Hyoshin Kim, Jungwon Kim, Jihee Kim, Sangyoon Park, Meeyoung Cha 3/20/2026

GeoReg: Weight-Constrained Few-Shot Regression for Socio-Economic Estimation using LLM

GeoReg uses LLMs with satellite imagery and geospatial data for socio-economic indicator estimation in data-scarce regions via few-shot regression.

Ax Zijian Liu 3/20/2026

Online Convex Optimization with Heavy Tails: Old Algorithms, New Regrets, and Applications

Research on Online Convex Optimization algorithms for heavy-tailed gradient distributions, extending beyond finite variance assumptions.

Ax Sepehr Maleki, Negar Pourmoazemi 3/20/2026

Pi-transformer: A prior-informed dual-attention model for multivariate time-series anomaly detection

Transformer architecture with dual attention for multivariate time-series anomaly detection using temporal invariants.

Ax Pooneh Mousavi, Lovenya Jain, Mirco Ravanelli, Cem Subakan 3/20/2026

Investigating Faithfulness in Large Audio Language Models

Framework evaluating faithfulness of chain-of-thought reasoning in large audio language models for multimodal tasks.

Ax Elaheh Akbari, Shansita Sharma, Ping He, Ahmadreza Moradipari, Kyungtae Han, Hamed Pirsiavash, Yikun Bai, Soheil Kolouri 3/20/2026

OT-MeanFlow3D: Bridging Optimal Transport and Meanflow for Efficient 3D Point Cloud Generation

Flow-matching models for 3D point cloud generation using optimal transport and meanflow for single-step inference acceleration.

Ax Ange-Cl\'ement Akazan, Verlon Roel Mbingui 3/20/2026

Splines-Based Feature Importance in Kolmogorov-Arnold Networks: A Framework for Supervised Tabular Data Dimensionality Reduction

KAN-based feature selection framework for tabular data via spline-based importance scoring. Specialized ML technique.

Ax Maryam Aliakbarpour, Vladimir Braverman, Junze Yin, Haochen Zhang 3/20/2026

Support Basis: Fast Attention Beyond Bounded Entries

Sub-quadratic attention algorithm removing bounded-entry restrictions for LLM inference speedup. Foundational LLM efficiency research.

Ax Seunghyeon Kim, Taesun Yeom, Jinho Kim, Wonpyo Park, Kyuyeun Kim, Jaeho Lee 3/20/2026

Activation Quantization of Vision Encoders Needs Prefixing Registers

Quantization technique for vision encoders using prefix registers to handle outliers. Optimization research for multimodal models.

Ax Ziyue Wang, Yayati Jadhav, Peter Pak, Amir Barati Farimani 3/20/2026

Image2Gcode: Image-to-G-code Generation for Additive Manufacturing Using Diffusion-Transformer Model

Diffusion-Transformer model converting images directly to G-code for 3D printing. Applied ML, domain-specific.

Ax Giulia Lanzillotta, Damiano Meier, Thomas Hofmann 3/20/2026

Heads collapse, features stay: Why Replay needs big buffers

Continual learning research on replay buffer size impact on feature retention vs. classifier forgetting. Specialized ML theory.

Ax Yifan Zhang, Wei Bi, Kechi Zhang, Dongming Jin, Jie Fu, Zhi Jin 3/20/2026

Weights to Code: Extracting Interpretable Algorithms from the Discrete Transformer

Algorithm extraction from Discrete Transformers via symbolic program synthesis. Addresses representation entanglement in interpretability.

Ax Injin Kong, Hyoungjoon Lee, Yohan Jo 3/20/2026

Mechanism Shift During Post-training from Autoregressive to Masked Diffusion Language Models

Research analyzing mechanistic changes when post-training autoregressive models into masked diffusion models. Studies model internals via circuit analysis.

Ax Qinglun Li, Anke Tang, Miao Zhang, Mengzhu Wang, Quanjun Yin, Li Shen 3/20/2026

A Unified Generalization Framework for Model Merging: Trade-offs, Non-Linearity, and Scaling Laws

Unified theoretical framework for model merging explaining effectiveness across heterogeneous fine-tuning hyperparameters with scaling laws.

Ax Rebecca Pelke, Joel Klein, Jose Cubero-Cascante, Nils Bosbach, Jan Moritz Joseph, Rainer Leupers 3/20/2026

Mixed-Precision Training and Compilation for RRAM-based Computing-in-Memory Accelerators

Mixed-precision training and compilation techniques for RRAM-based computing-in-memory ML accelerators with low bit-width constraints.

Ax Jingkun Liu, Yisong Yue, Max Welling, Yue Song 3/20/2026

Krause Synchronization Transformers

Krause Attention: principled attention mechanism addressing representation collapse and attention sink issues in transformers.

Ax Nicolas Zumarraga, Thomas Kaar, Ning Wang, Maxwell A. Xu, Max Rosenblattl, Markus Kreft, Kevin O'Sullivan, Paul Schmiedmayer, Patrick Langer, Robert Jakob 3/20/2026

TS-Haystack: A Multi-Scale Retrieval Benchmark for Time Series Language Models

Multi-scale retrieval benchmark for time series language models addressing long-context temporal localization under computational constraints.

Ax Shruti Joshi, Aaron Mueller, David Klindt, Wieland Brendel, Patrik Reizinger, Dhanya Sridhar 3/20/2026

Causality is Key for Interpretability Claims to Generalise

Position paper on causal inference requirements for valid and generalizable interpretability claims in LLM research.

Ax Ali Saheb Pasand, Johan Obando-Ceron, Aaron Courville, Pouya Bashivan, Pablo Samuel Castro 3/20/2026

Stable Deep Reinforcement Learning via Isotropic Gaussian Representations

Deep reinforcement learning stability improvement through isotropic Gaussian embeddings under non-stationary training dynamics.

Ax Hung-Hsuan Chen 3/20/2026

CeRA: Breaking the Linear Ceiling of Low-Rank Adaptation via Manifold Expansion

CeRA: improved parameter-efficient fine-tuning method that surpasses LoRA's linear constraints via manifold expansion with gating and dropout.

Ax Yongzhong Xu 3/20/2026

Optimizer-Induced Low-Dimensional Drift and Transverse Dynamics in Transformer Training

Analysis of transformer training trajectories under AdamW showing low-dimensional drift directions and batch-gradient alignment patterns.

Ax Daniel S. Berman, Brian Merritt, Stanley Ta, Dana Udwin, Amanda Ernlund, Jeremy Ratcliff, Vijay Narayan 3/20/2026

What You Read is What You Classify: Highlighting Attributions to Text and Text-Like Inputs

Explainable AI method for highlighting token attributions in text classification using transformers.

Ax Nilesh Jain, Rohit Yadav, Sagar Kotian, Claude AI 3/20/2026

AutoResearch-RL: Perpetual Self-Evaluating Reinforcement Learning Agents for Autonomous Neural Architecture Discovery

Framework for autonomous neural architecture and hyperparameter search using self-evaluating RL agents without human supervision.

Ax Angad Singh Ahuja 3/20/2026

Adversarial Latent-State Training for Robust Policies in Partially Observable Domains

Research on robust policy training in partially observable reinforcement learning under adversarial latent state distribution shifts.

Ax Xiangwen Wang, Ananth Balashankar, Varun Chandrasekaran 3/20/2026

Systematic Scaling Analysis of Jailbreak Attacks in Large Language Models

Systematic study of jailbreak attack scaling laws across LLM methods and model families using compute-bounded optimization framework.

Ax Haihua Luo, Xuming Ran, Tommi K\"arkk\"ainen, Huiyan Xue, Zhonghua Chen, Qi Xu, Fengyu Cong 3/20/2026

Representation Finetuning for Continual Learning

Research on parameter-efficient fine-tuning for continual learning using representation-level optimization instead of weight-level black-box methods.

Ax Wanyin Wu, Kanxue Li, Baosheng Yu, Haoyun Zhao, Yibing Zhan, Dapeng Tao, Hua Jin 3/20/2026

PREBA: Surgical Duration Prediction via PCA-Weighted Retrieval-Augmented LLMs and Bayesian Averaging Aggregation

Zero-shot surgical duration prediction combining retrieval-augmented LLMs with Bayesian averaging for resource management.

Ax Zakia Zaman, Praveen Gauravaram, Mahbub Hassan, Sanjay Jha, Wen Hu 3/20/2026

Privacy-Preserving Machine Learning for IoT: A Cross-Paradigm Survey and Future Roadmap

Survey of privacy-preserving machine learning mechanisms for IoT devices covering federated learning and edge computing approaches.

Ax Yongzhong Xu 3/20/2026

Spectral Edge Dynamics of Training Trajectories: Signal--Noise Geometry Across Scales

Analysis of transformer training dynamics via Spectral Edge Dynamics, identifying coherent optimization directions vs stochastic noise.

Ax Ting Gao, Stavros Orfanoudakis, Nan Lin, Elvin Isufi, Winnie Daamen, Serge Hoogendoorn 3/20/2026

Flow Matching Policy with Entropy Regularization

Diffusion-based reinforcement learning policy using flow matching with direct entropy regularization and efficient gradient computation.

Ax Ji-Hun Oh, Kianoush Falahkheirkhah, John Cheville, Rohit Bhargava 3/20/2026

Hallucination Detection in Virtually-Stained Histology: A Latent Space Baseline

Detects hallucinations in virtually-stained histology using latent space analysis and neural precursor method.

Ax Edric Tam, Barbara E Engelhardt 3/20/2026

Assessing the Distributional Fidelity of Synthetic Chest X-rays using the Embedded Characteristic Score

Method for evaluating synthetic chest X-ray quality using embedded characteristic scores.

Ax Harrish Thasarathan, Julian Forsyth, Thomas Fel, Matthew Kowal, Konstantinos G. Derpanis 3/20/2026

Universal Sparse Autoencoders: Interpretable Cross-Model Concept Alignment

Universal sparse autoencoders for discovering and aligning interpretable concepts across multiple neural networks.

Ax Anastasia N. Krouglova, Hayden R. Johnson, Basile Confavreux, Michael Deistler, Pedro J. Gon\c{c}alves 3/20/2026

Multifidelity Simulation-based Inference for Computationally Expensive Simulators

Multifidelity simulation-based inference framework for parameter estimation with expensive simulators.

Ax Alexandru Apostu, Silviu Gheorghe, Andrei H\^iji, Nicolae Cleju, Andrei P\u{a}tra\c{s}cu, Cristian Rusu, Radu Ionescu, Paul Irofti 3/20/2026

Detecting and Mitigating DDoS Attacks with AI: A Survey

Survey of AI-based methods for detecting and mitigating distributed denial-of-service attacks.

Ax Lovedeep Gondara, Jonathan Simkin, Shebnum Devji, Gregory Arbour, Raymond Ng 3/20/2026

ELM: A Hybrid Ensemble of Language Models for Automated Tumor Group Classification in Population-Based Cancer Registries

Ensemble of language models for automated tumor classification in cancer registry pathology reports.

Ax Adel Chehade, Edoardo Ragusa, Paolo Gastaldo, Rodolfo Zunino 3/20/2026

Hardware-Aware Neural Architecture Search for Encrypted Traffic Classification on Resource-Constrained Devices

Hardware-aware neural architecture search for encrypted traffic classification on IoT edge devices.

Ax Jacqueline Maasch, John Kalantari, Kia Khezeli 3/20/2026

CausalARC: Abstract Reasoning with Causal World Models

Testbed for evaluating AI reasoning with causal world models in low-data and out-of-distribution settings.

Ax Nikita Kornilov, David Li, Tikhon Mavrin, Aleksei Leonov, Nikita Gushchin, Evgeny Burnaev, Iaroslav Koshelev, Alexander Korotin 3/20/2026

Universal Inverse Distillation for Matching Models with Real-Data Supervision (No GANs)

Universal distillation method for training efficient one-step generators from diffusion and flow models without GANs.

Ax Yuanzhi Zhu, Xi Wang, St\'ephane Lathuili\`ere, Vicky Kalogeiton 3/20/2026

Soft-Di[M]O: Improving One-Step Discrete Image Generation with Soft Embeddings

Improves one-step image generation from masked diffusion models using soft embeddings to enable gradient flow for fine-tuning.

Ax Seyma Yaman Kayadibi 3/20/2026

Redundancy-as-Masking: Formalizing the Artificial Age Score (AAS) to Model Memory Aging in Generative AI

Artificial Age Score framework modeling memory aging patterns in large language models across conversational contexts.

Ax Gaoxiang Huang, Songning Lai, Yutao Yue 3/20/2026

Towards more holistic interpretability: A lightweight disentangled Concept Bottleneck Model

Lightweight disentangled concept bottleneck model for improved interpretability in neural networks.

Ax Zijian Liu 3/20/2026

Clipped Gradient Methods for Nonsmooth Convex Optimization under Heavy-Tailed Noise: A Refined Analysis

Theoretical analysis of clipped gradient optimization under heavy-tailed noise with refined convergence bounds.

Ax Jos\'e Pulido, Francesc Wilhelmi, Sergio Fortes, Alfonso Fern\'andez-Dur\'an, Lorenzo Galati Giordano, Raquel Barco 3/20/2026

Studying the Role of Synthetic Data for Machine Learning-based Wireless Networks Traffic Forecasting

Novel synthetic data generation method for wireless network traffic forecasting to augment training datasets.

Ax Marc Salvad\'o-Benasco, Aymane Kssim, Alexander Heinlein, Rolf Krause, Serge Gratton, Alena Kopani\v{c}\'akov\'a 3/20/2026

Multi-Preconditioned LBFGS for Training Finite-Basis PINNs

Multi-preconditioned LBFGS algorithm for training physics-informed neural networks using domain decomposition.

Ax Zanlin Ni, Shenzhi Wang, Yang Yue, Tianyu Yu, Weilin Zhao, Yeguo Hua, Tianyi Chen, Jun Song, Cheng Yu, Bo Zheng, Gao Huang 3/20/2026

The Flexibility Trap: Why Arbitrary Order Limits Reasoning Potential in Diffusion Language Models

Study showing arbitrary token generation order in diffusion LLMs doesn't improve reasoning despite flexibility.

Ax Yunwei Bai, Ying Kiat Tan, Yao Shu, Tsuhan Chen 3/20/2026

1S-DAug: One-Shot Data Augmentation for Robust Few-Shot Generalization

One-shot data augmentation technique for few-shot learning combining geometric perturbations with noise injection.