Robuta

https://openreview.net/forum?id=BySRH6CpW¬eId=BySRH6CpW Learning Discrete Weights Using the Local Reparameterization Trick | OpenReview Training binary/ternary networks using local reparameterization with the CLT approximation the local reparameterization trick learning discrete weights https://arxiv.org/html/2402.17824v1 Quantum glasses, reparameterization invariance, Sachdev-Ye-Kitaev models quantum glasses reparameterization invariance sachdev https://openreview.net/forum?id=2HFmicB8kh&referrer=%5Bthe%20profile%20of%20Vaibhav%20Seth%5D(%2Fprofile%3Fid%3D~Vaibhav_Seth1) Robust and Efficient Fine-tuning of LLMs with Bayesian Reparameterization of Low-Rank Adaptation |... Large Language Models (LLMs) are highly resource-intensive to fine-tune due to their enormous size. While low-rank adaptation is a prominent... https://openreview.net/forum?id=VMV8gefvq8 MCNC: Manifold-Constrained Reparameterization for Neural Compression | OpenReview The outstanding performance of large foundational models across diverse tasks, from computer vision to speech and natural language processing, has... mcnc manifold constrained reparameterization neural https://openreview.net/forum?id=KRVnpTbx7R DiVeQ: Differentiable Vector Quantization Using the Reparameterization Trick | OpenReview Vector quantization is common in deep models, yet its hard assignments block gradients and hinder end-to-end training. We propose DiVeQ, which treats... vector quantization reparameterization trick differentiable using openreview https://openreview.net/forum?id=WV1ZXTH0OIn Bayesian Optimization over Discrete and Mixed Spaces via Probabilistic Reparameterization |... We propose a theoretically-grounded method for Bayesian optimization over discrete and mixed search spaces and demonstrate state-of-the-art performance on a... bayesian optimization discrete mixed spaces via https://openreview.net/forum?id=zd2chYqgUj Generative Neural Reparameterization for Differentiable PDE-Constrained Optimization | OpenReview Partial-differential-equation (PDE)-constrained optimization is a well-worn technique for acquiring optimal parameters of systems governed by PDEs. However,... pde constrained optimization generative neural reparameterization differentiable https://openreview.net/forum?id=HkgxW0EYDS Scalable Model Compression by Entropy Penalized Reparameterization | OpenReview An end-to-end trainable model compression method optimizing accuracy jointly with the expected model size. model compression scalable entropy reparameterization openreview