Explainable Machine Learning in Finance - When Machines Surpass Human Intelligence
Explore causal and counterfactual explainability in AI/ML models for finance, addressing challenges when machines surpass human intelligence in complex, high-risk applications.
Operational Research for Fairness, Privacy and Interpretability in Machine Learning
Explore responsible AI through operational research, focusing on fairness, interpretability, and privacy in machine learning. Learn to leverage Integer Linear Programming for optimal, fair rule lists.
Do Androids Know They're Only Dreaming of Electric Sheep?
Explore transformer language models' internal representations to detect and analyze hallucinatory behavior in AI-generated content, with insights on task-dependent patterns and evaluation methods.
Deep Learning Interpretability and Explainability of Speech in a Clinical Context
Explore deep learning interpretability and explainability of speech in clinical settings, focusing on advanced techniques and their practical applications.
Frontiers of Foundation Models for Time Series
Explore foundation models for time series data, addressing challenges in science applications and discussing future research directions in deep learning for time series analysis.
Seeing Like a Language Model - Understanding Embeddings and Meaning in AI
Explore how language models process text and represent meaning through embeddings. Gain insights into building better retrieval-augmented LLM systems and designing natural interfaces for reading and writing.
Kolmogorov-Arnold Networks - Alternatives to Multi-Layer Perceptrons
Explore Kolmogorov-Arnold Networks as alternatives to MLPs, offering improved accuracy and interpretability in data fitting and PDE solving. Discover their potential for scientific collaboration and law discovery.
Interpreting Deep Neural Networks Towards Trustworthiness
Explore methods for interpreting neural networks, including contextual decomposition and adaptive wavelet distillation, to enhance model trustworthiness and performance.
Rapture of the Deep: Highs and Lows of Sparsity in Neural Networks
Explore sparsity in neural networks, its role in interpretability, and innovative optimization techniques beyond gradient descent for learning fast transforms.
Interpretable and Structure-Preserving Data-Driven Methods for Physical Simulations
Explore interpretable, structure-preserving data-driven methods for physical simulations. Discover techniques like DMD, LaSDI, and ROMs to accelerate and improve accuracy in various scientific applications.
Towards Monosemanticity: Decomposing Language Models with Dictionary Learning
Explore monosemanticity in neural networks through sparse autoencoders. Learn how extracting interpretable features enhances understanding of language model behavior and improves reasoning capabilities.