paper-with-me

홈 › Papers

Leveraging free energy in pretraining model selection for improved fine-tuning

2024-10-08 · Michael Munn, Susan Wei

Recent advances in artificial intelligence have been fueled by the development of foundation models such as BERT, GPT, T5, and Vision Transformers. These models are first pretrained on vast and diverse datasets and then adapted to specific downstream tasks, often with significantly less data. However, the mechanisms behind the success of this ubiquitous pretrain-then-adapt paradigm remain underexplored, particularly the characteristics of pretraining checkpoints that lend themselves to good downstream adaptation. We introduce a Bayesian model selection criterion, called the downstream free energy, which quantifies a checkpoint's adaptability by measuring the concentration of nearby favorable parameters for the downstream task. We demonstrate that this free energy criterion can be effectively implemented without access to the downstream data or prior knowledge of the downstream task. Furthermore, we provide empirical evidence that the free energy criterion reliably correlates with improved fine-tuning performance, offering a principled approach to predicting model adaptability.

📄 PDF Abstract BibTeX arXiv:2410.05612

Code (0)

등록된 구현이 없습니다.

Tasks

Model Selection

Methods 이 논문이 사용한 방법론

Refunds@Expedia|||How do I get a full refund from Expedia? “How do I get a full refund from Expedia? How do I get a full refund from Expedia? – Call ☎️ +1-(888) 829 (0881) or +1-805-330-4056 or +1-805-330-4056 for Quick Help &…
Attention 설명 없음
Gated Linear Unit A Gated Linear Unit, or GLU computes: $$ \mathrm{GLU}(a, b) = a \otimes \sigma(b) $$ It is used in natural language processing architectures, for example the Gated CNN,…
WordPiece 설명 없음
SentencePiece 설명 없음
Inverse Square Root Schedule Inverse Square Root is a learning rate schedule 1 / $\sqrt{\max\left(n, k\right)}$ where $n$ is the current training iteration and $k$ is the number of warm-up steps. This…
Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Cosine Annealing Cosine Annealing is a type of learning rate schedule that has the effect of starting with a large learning rate that is relatively rapidly decreased to a minimum value before…

Similar Papers 제목 키워드 기반

Leveraging Gauge Freedom for Learning Non-Gradient Population Dynamics of Stochastic Systems

2026-05-24 · Jules Berman, Tobias Blickhan, Benjamin Peherstorfer arxiv

Existing work on population dynamics inference often focuses on flows arising from vector fields that are the gradients of scalar potentials. Among all admissible flows that are compatible with the population dynamics, g…

Free-Energy-Gated Plasticity for Real-Time Online Motor Learning in Physical Human--Robot Interaction

2026-08-24 · Hiroki Sawada, Jun Tani arxiv

Fully online embodied learning requires synaptic adaptation to acquire new behaviors while preserving previously learned dynamics during ongoing interaction. We extend the Predictive-Coding-inspired Variational Recurrent…

Molecular Geometry Pretraining with SE(3)-Invariant Denoising Distance Matching

2022-06-27 · Shengchao Liu, Hongyu Guo, Jian Tang

Molecular representation pretraining is critical in various applications for drug and material discovery due to the limited number of labeled molecules, and most existing work focuses on pretraining on 2D molecular graph…

Denoisingmolecular representation

Farewell to Aimless Large-scale Pretraining: Influential Subset Selection for Language Model

2023-05-22 · Xiao Wang, Weikang Zhou, Qi Zhang, Jie zhou 외

Pretrained language models have achieved remarkable success in various natural language processing tasks. However, pretraining has recently shifted toward larger models and larger data, and this has resulted in significa…

Language ModelingLanguage Modelling

Is It a Free Lunch for Removing Outliers during Pretraining?

2024-02-19 · Baohao Liao, Christof Monz

With the growing size of large language models, the role of quantization becomes increasingly significant. However, outliers present in weights or activations notably influence the performance of quantized models. Recent…

Quantization