Michael Poli

23 papers · 2020–2025 · 5 conferences · across top CS/AI conferences

Achievements

🐣 Hot Topic Early Bird 🌉 Interdisciplinary Bridge 🧭 Keyword Pioneer 🌍 Conference Polyglot (5) 🏃 Academic Marathon (5) 🐝 Cross-Pollinator (6) 🌈 Renaissance Researcher (7) 🗺️ Taxonomy Completionist (34) 👑 Triple Crown 🤝 Dynamic Duo (17) ❓ The Questioner 💎 Century Club (23) 🔥 Unstoppable (6) 🗃️ Keyword Collector (96) ⚡ Prolific Year (5)

Conferences

NIPS (10) ICML (6) ICLR (5) AISTATS (1) IJCAI (1)

Top co-authors

Stefano Massaroli (17) Christopher Re (10) Stefano Ermon (9) Jinkyoo Park (8) Atsushi Yamashita (6) Hajime Asama (5) Tri Dao (4) Rom Parnichkun (4) Taiji Suzuki (4) Atri Rudra (4)

Keywords

neural network (3) dynamical system (3) deep learning (2) neural differential equation (2) neural ordinary differential equation (2) attention mechanism (2) model architecture (2) language modeling (2) model compression (2) knowledge distillation (1) spatio-temporal modeling (1) image classification (1) latent state space (1) time series generation (1) image compression (1) fourier transform (1) transfer learning (1) optimal control (1) time series classification (1) computational efficiency (1)

Papers

STAR: Synthesis of Tailored Architectures ICLR 2025

Quantifying Memory Utilization with Effective State-Size ICML 2025

Mechanistic Design and Scaling of Hybrid Architectures ICML 2024

Zoology: Measuring and Improving Recall in Efficient Language Models ICLR 2024

State-Free Inference of State-Space Models: The *Transfer Function* Approach ICML 2024

Effectively Modeling Time Series with Simple Discrete State Spaces ICLR 2023

Deep Latent State Space Models for Time-Series Generation ICML 2023

Hyena Hierarchy: Towards Larger Convolutional Language Models ICML 2023

Learning Efficient Surrogate Dynamic Models with Graph Spline Networks NIPS 2023

Monarch Mixer: A Simple Sub-Quadratic GEMM-Based Architecture NIPS 2023

Ideal Abstractions for Decision-Focused Learning AISTATS 2023

Laughing Hyena Distillery: Extracting Compact Recurrences From Convolutions NIPS 2023

HyenaDNA: Long-Range Genomic Sequence Modeling at Single Nucleotide Resolution NIPS 2023

Transform Once: Efficient Operator Learning in Frequency Domain NIPS 2022

Neural Solvers for Fast and Accurate Numerical Optimal Control ICLR 2022

Which Shortcut Cues Will DNNs Choose? A Study from the Parameter-Space Perspective ICLR 2022

Monarch: Expressive Structured Matrices for Efficient and Accurate Training ICML 2022

Self-Similarity Priors: Neural Collages as Differentiable Fractal Representations NIPS 2022

Differentiable Multiple Shooting Layers NIPS 2021

Neural Hybrid Automata: Learning Dynamics With Multiple Modes and Stochastic Transitions NIPS 2021

WATTNet: Learning to Trade FX via Hierarchical Spatio-Temporal Representation of Highly Multivariate Time Series IJCAI 2020

Hypersolvers: Toward Fast Continuous-Depth Models NIPS 2020

Dissecting Neural ODEs NIPS 2020