PROGRAM
Days: Wednesday, September 25th Thursday, September 26th Friday, September 27th
Wednesday, September 25th
View this program: with abstractssession overviewtalk overview
08:30-10:10 Session 1A: Learning and Networks I
Invited Session
Chairs:
Bruce Hajek and R Srikant
Location: Auditorium
| 08:30 | Opportunistic Learning for Markov Decision Systems with Application to Smart Robots (abstract) |
| 08:50 | Reinforcement Learning from Human Feedback without Reward Inference (abstract) |
| 09:10 | Sharp information-theoretic thresholds for shuffled linear regression (abstract) |
| 09:30 | Transform Method for Markov-Modulated Queues (abstract) |
| 09:50 | Caching or Pre-fetching? The Role of Hazard Rates (abstract) |
08:30-10:10 Session 1B: Sequential Methods I
Invited Session
Chairs:
Location: 1040
| 08:30 | Sequential Design of Experiments in Causal Systems (abstract) |
| 08:50 | Robust Score-Based Quickest Change Detection (abstract) |
| 09:10 | Minimax Rate Optimal Algorithms For High-Dimensional Stochastic Linear Bandits (abstract) |
| 09:30 | Higher-criticism for multi-sensor sparse change-point detection (abstract) |
08:30-10:10 Session 1C: Optimization I
Invited Session
Chairs:
Location: 2000N
| 08:30 | Distributed Difference of Convex Optimization (abstract) |
| 08:50 | Closing the Duality Gap on Integer Linear Programming (abstract) |
| 09:10 | Optimal Regularization for a Data Source (abstract) |
| 09:30 | Sharply predicting optimization trajectories in high-dimensions: State evolutions beyond first-order algorithms (abstract) |
| 10:00 | On the scalability of large graph methods for kernel-based machine learnin (abstract) |
08:30-10:10 Session 1D: Contraction, Entropy and Related Topics I
Invited Session
Chair:
Location: 3000
| 08:30 | Nonlinear observers with online error bounds (abstract) |
| 08:50 | Control with minimum data-transmission (abstract) |
| 09:10 | Lyapunov-like converse result for strong forward invariance (abstract) |
| 09:30 | On topological entropy of switched cascade systems (abstract) |
08:30-10:10 Session 1E: Information Theory, Statistics, and Data Science I
Invited Session
Chairs:
Location: 1030
| 08:30 | Theoretical Guarantees of Data Augmented Last Layer Retraining Methods (abstract) |
| 08:50 | Differential Privacy in the Large Composition Regime (abstract) |
| 09:10 | Optimal Redundancy in Exact Channel Synthesis (abstract) |
| 09:30 | The Power Method with Row Erasures (abstract) |
| 09:50 | Robust Distribution Estimation via Partial Optimal Transport (abstract) |
10:30-12:10 Session 2A: Learning and Networks II
Invited Session
Chairs:
R Srikant and Bruce Hajek
Location: Auditorium
| 10:30 | Self-Play Preference Optimization for Language Model Alignment (abstract) |
| 10:50 | Pluralistic Alignment Framework for Learning from Heterogeneous Preferences (abstract) |
| 11:10 | Shaping opinions in social networks with shadow banning (abstract) |
| 11:30 | Prelimit Coupling and Steady-State Convergence of Constant-stepsize Nonsmooth Contractive Stochastic Approximation (abstract) |
| 11:50 | Beyond PCA: A Gram-Schmidt Approach to Feature Extraction (abstract) PRESENTER: Bahram Yaghooti |
10:30-12:10 Session 2B: Quantum and Coding Techniques
Location: 2000N
| 10:30 | Acceptance in Postselected Asymptotic Quantum State Discrimination (abstract) |
| 10:50 | Solving the B-SAT Problem Using Quantum Computing: Smaller is Sometimes Better (abstract)  PRESENTER: Ahmad Bennakhi |
| 11:10 | Authenticated partial correction over AV-MACs: toward characterization and coding (abstract) |
| 11:30 | Graph Codes for Dual-Parameter Barrier Channels (abstract) PRESENTER: Saar Stern |
| 11:50 | Quantum Margulis Codes (abstract)  PRESENTER: Michele Pacenti |
10:30-12:10 Session 2C: Grid-Forming Inverter Modeling and Control I
Invited Session
Chair:
Location: 1030
| 10:30 | Grid-forming control of renewable generation and hybrid ac/dc power systems (abstract) |
| 10:50 | Equivalent-circuit Models for Grid-forming Inverters under Unbalanced Steady-State Operating Conditions (abstract) |
| 11:10 | Hierarchical Control of Distribution Grids with Application to Transmission-Distribution Fast Frequency Control (abstract) |
| 11:30 | Sensitivity to Hopf Bifurcation in Grid-Following Inverters Using Normal Vector Methods (abstract) |
| 11:50 | Equivalent-circuit Modeling Framework to Facilitate Large-signal Stability Assessment of Inverter-based Resources (abstract) |
10:30-12:10 Session 2D: Contraction, Entropy and Related Topics II
Invited Session
Chair:
Location: 1040
| 10:30 | Minimum Data-Rate for Emulating a Linear Feedback System (abstract) |
| 10:50 | A sufficient condition for 2-contraction of a feedback interconnection (abstract) |
| 11:10 | Remote robust state estimation problem (abstract) |
| 11:30 | Reachability Analysis of Dynamical Systems: A Mixed Monotone Contracting Approach (abstract) |
10:30-12:10 Session 2E: Channel Coding Techniques
Location: 3000
| 10:30 | SPIR with Colluding and Non-Replicated Servers from a Noisy Channel (abstract) PRESENTER: Amirhossein Shekofteh |
| 10:50 | Strategies for Rate Optimization in Joint Sensing and Communication over Channels with Memory (abstract) |
| 11:10 | Contraction Coefficients of Product Symmetric Channels (abstract) PRESENTER: Dongmin Lee |
| 11:30 | Finite-Length Coding Bounds via Guesswork (abstract) PRESENTER: Alexander Mariona |
| 11:50 | Higher-order Interpretations of Deepcode, a Learned Feedback Code (abstract) |
14:30-16:30 Session 3A: Quantum Algorithms and Applications
Invited Session
Chairs:
Location: 1030
| 14:30 | Feedback-based quantum algorithms: Extensions to real-time feedback (abstract) |
| 14:50 | Specialized Numerical Optimization Algorithms for QIS (abstract) |
| 15:10 | Transform Arbitrary Good Quantum LDPC Codes into Good Geometrically Local Codes in Any Dimension (abstract) |
| 15:30 | CANOE: The Classically Assisted Non-Orthogonal Eigensolver (abstract) |
| 15:50 | New methods for quantum circuit synthesis with symmetry-respecting gates (abstract) |
| 16:10 | On Non-Local Coset Guessing Quantum Games (abstract) |
14:30-16:30 Session 3B: Duality in Control, Inference, and Learning
Invited Session
Chairs:
Location: 1040
| 14:30 | Weyl Calculus and Exactly Solvable Schrödinger Bridges with Quadratic State Cost (abstract) PRESENTER: Abhishek Halder |
| 14:50 | Stochastic interpolants: A unifying framework for flows and diffusions (abstract) |
| 15:10 | A Fisher-Rao gradient flow for entropy-regularised Markov decision processes in Polish spaces (abstract) |
| 15:30 | Nonlinear Filtering with Brenier Optimal Transport Maps (abstract) |
| 15:50 | Scalable computations for nonlinear balanced truncation model reduction (abstract) |
| 16:10 | Harmonic Path Integral Diffusion (abstract) |
14:30-16:30 Session 3C: Distributed Control, Optimization, and Learning I
Invited Session
Chairs:
Location: Auditorium
| 14:30 | Solving Markov Decision Processes via Reward Balancing (abstract) |
| 14:50 | Markov Chain Variance Estimation: A Stochastic Approximation Approach (abstract) |
| 15:10 | Toward Parameter-free Decentralized Optimization (abstract) |
| 15:30 | Localized spectral representations for reinforcement learning in networked MDPs (abstract) |
| 15:50 | Classification with a Network of Partially Informative Agents: Enabling Wise Crowds from Individually Myopic Classifiers (abstract) |
14:30-16:30 Session 3D: Control Design Techniques
Location: 3000
| 14:30 | Generalized Barrier Functions: Integral Conditions & Recurrent Relaxations (abstract) |
| 14:50 | Adaptive Online Model Update Algorithm for Predictive Control in Networked Systems (abstract) |
| 15:10 | Optimal Infinite-Horizon Mixed $H_2 / H_\infty$ Control (abstract) PRESENTER: Taylan Kargin |
| 15:30 | A synthesis approach for distributed H2 control problems with communication delays (abstract) PRESENTER: Yuanji Zou |
14:30-16:30 Session 3E: Privacy and Deception
Location: 2000N
| 14:30 | A Hypothesis Testing-based Framework for Cyber Deception with Sludging (abstract) PRESENTER: Pramod K. Varshney |
| 14:50 | Privacy-Utility Tradeoff Based on $\alpha$-lift (abstract) |
| 15:10 | A New Framework for Designing Polynomial Codes for Private Information Retrieval (abstract) |
| 15:30 | Distributed Matrix Multiplication: Download Rate, Randomness and Privacy Trade-Offs (abstract) PRESENTER: Amirhosein Morteza |
| 15:50 | Influence Chain Recovery and Application in Distributed Fact-Checking (abstract) |
Thursday, September 26th
View this program: with abstractssession overviewtalk overview
08:30-10:10 Session 4A: Learning and Networks III
Invited Session
Chairs:
R Srikant and Bruce Hajek
Location: Auditorium
| 08:30 | Active Exploration for Minimax Estimation in MDPs (abstract) |
| 08:50 | Learning to Defer in Content Moderation: The Human-AI Interplay (abstract) |
| 09:10 | Minkowski Descent: An Algorithm for Stochastic Global Optimization (abstract) |
| 09:30 | On solving inverse problems in computer vision using latent diffusion models (abstract) |
| 09:50 | Quickest Change Detection Using Mismatched CUSUM (abstract) |
08:30-10:10 Session 4B: Information Theory, Statistics, and Data Science II
Invited Session
Chairs:
Location: 1040
| 08:30 | Differential Privacy and Sublinear Time are Incompatible Sometimes (abstract) |
| 08:50 | Differentially Private Space-Efficient Algorithms for Frequency Moment Estimation in the Turnstile Model (abstract) |
| 09:10 | Optimized Mechanisms for Differentially Private Composition (abstract) |
| 09:30 | Differentially Private Graph Diffusion with Applications in Personalized PageRanks (abstract) |
| 09:50 | Training Language Models on Private Federated Data in the Age of LLMs (abstract) |
08:30-10:10 Session 4C: Optimization II
Invited Session
Chairs:
Location: 1030
| 08:30 | Robustly Learning Single-Index Models via Alignment Sharpness (abstract) |
| 08:50 | Synthetic Principal Component Design: Fast Covariate Balancing with Synthetic Controls (abstract) |
| 09:10 | Stochastic-Gradient-based Interior-Point Methods (abstract) |
| 09:30 | Convergence of Gradient Descent with Small Initialization for Unregularized Matrix Completion (abstract) |
| 09:50 | Structured Regularization for SPD Optimization with Side Information (abstract) |
08:30-10:10 Session 4D: Capacity of Communication Channels
Location: 3000
| 08:30 | On the Capacity of Scalar Gaussian Channels Subject to State Obfuscation (abstract) PRESENTER: Matthew Ho |
| 08:50 | Capacity of a Complex-Valued Gaussian Channel with Frequency Offset (abstract) PRESENTER: Omkar Mujumdar |
| 09:10 | An Algorithm for Computing the Capacity of Symmetrized KL Information for Discrete Channels (abstract) |
| 09:30 | Understanding Neural Population Communication with Latent Channels (abstract) |
08:30-10:10 Session 4E: Sequential Methods II
Invited Session
Chairs:
Location: 2000N
| 08:30 | Non-Adaptive Multi-Stage Algorithm for Group Testing with Prior Statistics (abstract) PRESENTER: Alejandro Cohen |
| 08:50 | Post-Contextual Bandit Inference with Adaptive Stopping (abstract) |
| 09:10 | Power-one sequential tests, confidence sequences, e-processes and e-detectors (abstract) |
| 09:30 | Active Change Point Anomaly Detection Over Composite Hypotheses (abstract) |
| 09:50 | Private Online Community Detection for Censored Block Models (abstract) |
10:30-12:10 Session 5A: Games, Markets, and Learning
Invited Session
Chairs:
Location: Auditorium
| 10:30 | Incentive Analysis for Agent Participation in Federated Learning (abstract) |
| 10:50 | Instance-Optimality in Online Learning (abstract) |
| 11:10 | Incentives for Exploration at Market Equilibrium (abstract) |
| 11:30 | Tractable Equilibrium Computation in Markov Games through Risk Aversion (abstract) |
| 11:50 | Designing interventions in multi-activity network games for forest resource management (abstract) |
10:30-12:10 Session 5B: Quantum Error-Correction
Invited Session
Chairs:
Location: 1040
| 10:30 | Quantum LDPC Codes: Constructions and Decoders Handling Degeneracy (abstract) |
| 10:50 | Binary Triorthogonal and CSS-T Codes for Quantum Error Correction (abstract) |
| 11:10 | GNarsil: Splitting Stabilizers into Gauges (abstract) |
| 11:30 | Erasure Decoding for Quantum LDPC Codes via Belief Propagation with Guided Decimation (abstract) |
| 11:50 | Unified and Generalized Approach to Entanglement-Assisted Quantum Error Correction (abstract) |
10:30-12:10 Session 5C: Federated, Distributed, and Statistical Learning
Location: 2000N
| 10:30 | Communication-Efficient Federated Learning via Sparse Training with Regularized Error Correction (abstract) |
| 10:50 | Communication-efficient and Decentralized Federated Minimax Optimization (abstract) PRESENTER: Sajjad Ghiasvand |
| 11:10 | A Masked Pruning-Based Algorithm for Dimensionality Reduction in Federated Learning Systems (abstract) |
| 11:30 | Temporal Predictive Coding for Gradient Compression in Distributed Learning (abstract) |
| 11:50 | Learning the eye of the beholder: Statistical modeling and estimation for personalized color perception (abstract) |
10:30-12:10 Session 5D: Quantization, Compression, and Computation
Location: 3000
| 10:30 | Strategic Quantization with Quadratic Distortion Measures (abstract) |
| 10:50 | Lossless Binary Image Compression Using Learned Multi-level Dictionaries (abstract) PRESENTER: Renu Rameshan |
| 11:10 | Multi-Functional Distributed Computing (abstract) PRESENTER: Derya Malak |
10:30-12:10 Session 5E: Information Theory, Statistics, and Data Science III
Invited Session
Chairs:
Location: 1030
| 10:30 | Langevin Unlearning: A New Perspective of Noisy Gradient Descent for Machine Unlearning (abstract) |
| 10:50 | You Shall Not Leak Your Brother's Data: Privacy Control in the Presence of Correlation (abstract) |
| 11:10 | Restricted Converse Results for g-Group Complete-{s} Pliable Index Coding Problems (abstract) |
| 11:30 | Differentially Private Sinkhorn Algorithm (abstract) |
| 11:50 | Metric Clustering From Triplet Comparisons (abstract) |
16:00-17:00 Session 6: PLENARY
Location: Electrical and Computer Engineering Building (1002 ECEB)
Location: 1030
Friday, September 27th
View this program: with abstractssession overviewtalk overview
08:30-10:10 Session 7A: Electric Power Systems I
Invited Session
Chairs:
Location: 1030
| 08:30 | Braess’ Paradoxes in Electrified Transportation (abstract) |
| 08:50 | Information Structures in AC/DC Grids (abstract) |
| 09:10 | Fair-Over-Time Distributed Energy Resource Coordination (abstract) |
| 09:30 | Transferable Learning of GCN Sampling Graph Data Clusters from Different Power Systems (abstract) PRESENTER: Tong Wu |
| 09:50 | Acknowledgment Feedback in Directed Graphs via Deterministic and Probabilistic Strategies (abstract) |
08:30-10:10 Session 7B: Distributed Machine Learning
Invited Session
Chairs:
Location: Auditorium
| 08:30 | The Sample-Communication Complexity Trade-off in Federated Q-Learning (abstract) |
| 08:50 | Incentive Mechanism Design for Federated Learning with Unstateful Clients (abstract) |
| 09:10 | How Do System Dynamics Affect the Convergence of Federated Learning? (abstract) |
| 09:30 | collaborative exploration of unknown environments (abstract) |
| 09:50 | Personalized Federated Learning with Shared Linear Representation: Statistically Optimal Rates (abstract) |
08:30-10:10 Session 7C: Information Theory, Statistics, and Data Science IV
Invited Session
Chairs:
Location: 1040
| 08:30 | A robust testing approach to measuring model varability (abstract) |
| 08:50 | An Information Geometric Analysis of Noise Contrastive Estimation (abstract) PRESENTER: Hasan Sabri Melihcan Erol |
| 09:10 | Impossibility of latent inner product recovery via rate distortion (abstract) PRESENTER: Shenduo Zhang |
| 09:30 | OpAMP: Linear Regressions with Erasures (abstract) |
| 09:50 | Chernoff Information as a Privacy Constraint for Adversarial Classification (abstract) PRESENTER: Ayse Unsal |
08:30-10:10 Session 7D: Wireless Networks
Location: 3000
| 08:30 | Rendering Wireless Environments Useful for Gradient Estimators: A Zero-Order Stochastic Federated Learning Method (abstract) PRESENTER: Elissa Mhanna |
| 08:50 | Mobility in Age-Based Gossip Networks (abstract) |
| 09:10 | Linear-time Scheduling for Time-varying Wireless Networks via Randomization (abstract) |
| 09:30 | Traffic-driven Spectrum and Power Allocation via Scalable Multi-agent Reinforcement Learning (abstract) |
10:30-12:30 Session 8A: Robust Inference and Learning
Invited Session
Chairs:
Location: 1030
| 10:30 | Value Mirror Descent For Reinforcement Learning (abstract) |
| 10:50 | Robust Conditional Value-at-Risk Based Risk-Sensitive Reinforcement Learning (abstract) |
| 11:10 | Characterization of Slow Convergence in Word-of-Mouth Social Learning (abstract) |
| 11:30 | Adaptive Text Watermark for Large Language Models (abstract) |
| 11:50 | Achieving the Asymptotically Optimal Sample Complexity of Offline Reinforcement Learning: A DRO-Based Approach (abstract) |
| 12:10 | Distributionally Robust Optimal Transport for Improved Model Robustness (abstract) |
10:30-12:10 Session 8B: Electric Power Systems II
Invited Session
Chairs:
Location: Auditorium
| 10:30 | Balancing Fairness and Efficiency in Power Systems with Large and Small Users (abstract) |
| 10:50 | Decentralized Equitable Energy Access in Energy Communities (abstract) |
| 11:10 | Stability-Constrained Voltage Control in Distribution Grids (abstract) |
| 11:30 | Exploring Feasible Transition from Local to Global Optima in Power Systems (abstract) |
| 11:50 | Reverse Kron Reduction on Trees (abstract) |
10:30-12:30 Session 8C: Game Theory
Location: 2000N
| 10:30 | Linearly Solvable General-Sum Markov Games (abstract) |
| 10:50 | Constrained Correlated Equilibria (abstract) |
| 11:10 | Fairness of Equilibrium in Proportional Sharing Mechanisms with Bid Costs (abstract) |
| 11:30 | Price of Anarchy of a Censorship Policy removing Misinformation in a Social Network (abstract) |
| 11:50 | Suppressing Overestimation in Q-Learning through Adversarial Behaviors (abstract) PRESENTER: Hyeann Lee |
| 12:10 | Non-Monotone Variational Inequalities (abstract) |
10:30-12:10 Session 8D: Large Language Models
Invited Session
Chairs:
Bin Hu and Huan Zhang
Location: 3000
| 10:30 | From Large Language Models to Large Agent Models: Planning and Reasoning with Physical World Knowledge (abstract) |
| 10:50 | Mitigating Backdoor Threats to Large Language Models (abstract) |
| 11:10 | Foundation Models for Robotic Manipulation: Opportunities and Challenges (abstract) |
| 11:30 | Dependence Induced Representations (abstract) PRESENTER: Xiangxiang Xu |
| 11:50 | Efficient LLM Serving via Lossy Computation (abstract) |
10:30-12:10 Session 8E: Distributed Control, Optimization, and Learning II
Invited Session
Chairs:
Location: 1040
| 10:30 | On the role of network structure in learning to coordinate with bounded rationality (abstract) |
| 10:50 | Regularity and Approximations for Partially Observed Stochastic Control and Applications to Learning (abstract) |
| 11:10 | Locally Interdependent Multi-Agent MDP: Theoretical Framework for Decentralized Agents with Dynamic Dependencies (abstract) |
| 11:30 | Do LLM Agents Have Regret? A Case Study in Online Learning and Games (abstract) |