A Sober Look at LLMs for Material Discovery: Are They Actually Good for Bayesian Optimization Over Molecules?
The study focuses on the role of Large Language Models (LLMs) in accelerating Bayesian optimization in molecular…
Strong convexity-guided hyper-parameter optimization for flatter losses
The authors introduce a novel approach to hyper-parameter optimization in machine learning by focusing on the strong…
Simulated Overparameterization
The study presents Simulated Overparametrization (SOP), a novel paradigm that combines the efficiency of compact models with…
PAC Learnability under Explanation-Preserving Graph Perturbations
The article delves into the use of graph neural networks (GNN) in dealing with complex relationships and…
Federated Learning Can Find Friends That Are Beneficial
The paper presents a new algorithm for Federated Learning (FL) that identifies beneficial collaborations among clients. The…
Causal Representation Learning from Multiple Distributions: A General Setting
This paper discusses causal representation learning in a nonparametric setting, focusing on multiple distributions without assuming hard…
NITO: Neural Implicit Fields for Resolution-free Topology Optimization
The article presents Neural Implicit Topology Optimization (NITO), a deep learning approach for topology optimization in engineering…
On diffusion models for amortized inference: Benchmarking and improving stochastic control and sampling
The article studies the problem of training diffusion models to sample from a distribution with a given…
Hydragen: High-Throughput LLM Inference with Shared Prefixes
This research presents Hydragen, a hardware-aware model that improves the efficiency of transformer-based large language models (LLMs)…
Hydra: Sequentially-Dependent Draft Heads for Medusa Decoding
The article presents a novel method called ‘Hydra heads’ to improve the efficiency of speculative decoding in…