Papers by Aldeida Aleti
A Semantic-Aware Layer-Freezing Approach to Computation-Efficient Fine-Tuning of Language Models (2025.findings-acl)
Copied to clipboard
| Challenge: | Existing work on how to finetune but neglects the issue of where to fine-tune language models is expensive. |
| Approach: | They propose to use transition traces of latent representation to compute deviations (or loss) and then estimate the gain of each layer in reducing deviation (or gain). |
| Outcome: | The proposed approach outperforms baseline methods and is cost-benefit balanced. |
Beyond Neural Incompatibility: Cross-Scale Knowledge Transfer in Large Language Models through Latent Semantic Alignment (2026.findings-acl)
Copied to clipboard
| Challenge: | Existing methods that reuse layer parameters are limited by incompatibility . a central challenge is to make cross-scale knowledge transfer effective and efficient . |
| Approach: | They propose a method that uses latent semantic alignment to facilitate cross-scale knowledge transfer . they use activations to pair target and source layers in latent space to achieve alignment . |
| Outcome: | The proposed method is effective when source and target models differ in architecture and parameterization. |