Superposition-Transformer

Public

a novel architecture that leverages Autoencoders to superimpose the hidden representations of a base model and a fine-tuned model within a shared parameter space. Using B-spline-based blending coefficients and autoencoders that adaptively reconstruct the original hidden states based on the input data distribution.

artificial-intelligence autoencoders deep-learning language-model llm llm-training

Creat：2024-10-22T23:22:11

Update：2025-01-08T16:46:59

132

Stars

Stars Increase

Related projects

Tensorflow

deep-learning

An Open Source Machine Learning Framework for Everyone

191001

1年前

+37today

AutoGPT

AutoGPT is the vision of accessible AI for everyone, to use and to build on. Our mission is to provide the tools, so that you can focus on what matters.

177399

5个月前

+44today