Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Concept Videos

Linearization and Approximation01:26

Linearization and Approximation

154
Linearization is a mathematical technique used to approximate complex, nonlinear functions with simpler linear models in the vicinity of a chosen reference point. The method is based on the idea that, although a function may be difficult to evaluate exactly, its behavior near a specific input value can often be closely approximated by the tangent line at that point. This approach is particularly useful when small deviations from a known value are involved.Consider the square root function, for...
154
Residuals and Least-Squares Property01:11

Residuals and Least-Squares Property

9.8K
The vertical distance between the actual value of y and the estimated value of y. In other words, it measures the vertical distance between the actual data point and the predicted point on the line
If the observed data point lies above the line, the residual is positive, and the line underestimates the actual data value for y. If the observed data point lies below the line, the residual is negative, and the line overestimates the actual data value for y.
The process of fitting the best-fit...
9.8K
Application of Linearization and Approximation01:29

Application of Linearization and Approximation

132
A drone flying through complex terrain often relies on more than one sensing method to estimate small changes in altitude. Along with direct measurements, air pressure provides a useful indirect indicator of vertical movement. Atmospheric pressure decreases as altitude increases, and this relationship is commonly described using an exponential model. Although accurate, converting pressure measurements into altitude values requires calculations that are too complex to perform repeatedly during...
132
Reducing Line Loss01:18

Reducing Line Loss

440
In a three-phase circuit, line loss is an indicator of energy dissipated as heat due to the resistance of transmission lines. To address this, incorporating transformers into the system—a step-up transformer at the source and a step-down transformer at the load—is a strategic solution. Two three-phase transformers are introduced to improve this.
With a step-up transformer at the source, the voltage is increased, thereby reducing the current in the transmission lines since power loss in...
440
Calibration Curves: Linear Least Squares01:20

Calibration Curves: Linear Least Squares

5.1K
A calibration curve is a plot of the instrument's response against a series of known concentrations of a substance. This curve is used to set the instrument response levels, using the substance and its concentrations as standards. Alternatively, or additionally, an equation is fitted to the calibration curve plot and subsequently used to calculate the unknown concentrations of other samples reliably.
For data that follow a straight line, the standard method for fitting is the linear...
5.1K
Reconstruction of Signal using Interpolation01:10

Reconstruction of Signal using Interpolation

829
Signal processing techniques are essential for accurately converting continuous signals to digital formats and vice versa. When a continuous signal is sampled with a period T, the resulting sampled signal exhibits replicas of the original spectrum in the frequency domain, spaced at intervals equal to the sampling frequency. To handle this sampled signal, a zero-order hold method can be applied, which creates a piecewise constant signal by retaining each sample's value until the next...
829

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Unfractionated Heparin Attenuates Histone-Induced Pulmonary Endothelial Glycocalyx Injury: A Preliminary Study on the Roles of the ROCK Pathway and Heparanase‑1.

Journal of inflammation research·2026
Same author

Common clonal hematopoiesis driver mutations have disparate effects on macrophage cytokines, clonal expansion, and atherogenesis.

JCI insight·2025
Same author

PSOSP uncovers pervasive SOS-independent prophages with distinct genomic and host traits in bacterial genomes.

iMeta·2025
Same author

Long-term prognostic value of thyroid hormone levels in chronic critical illness patients.

Annals of medicine·2025
Same author

A Case Report of <i>Mycoplasma pneumoniae</i>-induced fulminant myocarditis in a 15-year-old male leading to cardiogenic shock and electrical storm.

Frontiers in cardiovascular medicine·2024
Same author

Risk factors of acute respiratory distress syndrome in sepsis caused by intra-abdominal infections: A retrospective study.

Surgery·2024

Related Experiment Video

Updated: Mar 24, 2026

Author Spotlight: Advancing 3D Cytoarchitecture Analysis - Rapid Volumetric Reconstruction of the Human Brain
06:52

Author Spotlight: Advancing 3D Cytoarchitecture Analysis - Rapid Volumetric Reconstruction of the Human Brain

Published on: January 26, 2024

3.0K

WLR: Well-conditioned linear reconstruction for retraining-free pruning of LLMs.

Siqi Li1, Jingyang Xiang1, Jiateng Wei1

  • 1Institute of Cyber-Systems and Control, Zhejiang University, Hangzhou, 310027, China.

Neural Networks : the Official Journal of the International Neural Network Society
|March 22, 2026
PubMed
Summary

Well-conditioned Linear Reconstruction (WLR) enhances large language models (LLMs) by efficiently compensating for structured pruning without adding parameters. This method improves numerical stability and model performance, surpassing existing techniques.

Keywords:
Deep neural networksLarge language modelModel compressionNetwork pruning

Related Experiment Videos

Last Updated: Mar 24, 2026

Author Spotlight: Advancing 3D Cytoarchitecture Analysis - Rapid Volumetric Reconstruction of the Human Brain
06:52

Author Spotlight: Advancing 3D Cytoarchitecture Analysis - Rapid Volumetric Reconstruction of the Human Brain

Published on: January 26, 2024

3.0K

Area of Science:

  • Artificial Intelligence
  • Machine Learning
  • Natural Language Processing

Background:

  • Structured pruning reduces large language model (LLM) size and computational needs.
  • Current compensation methods for LLM pruning often introduce extra parameters or overlook numerical instability.
  • Ill-conditioning in LLMs can lead to inefficient and unstable optimization solutions.

Purpose of the Study:

  • To introduce a novel structured pruning compensation method that avoids additional parameters.
  • To address and mitigate the ill-conditioning problem in LLMs during pruning compensation.
  • To improve the efficiency and numerical stability of pruned LLMs.

Main Methods:

  • Proposed Well-conditioned Linear Reconstruction (WLR) for structured pruning compensation.
  • WLR reconstructs pruned layers via a linear combination of preserved channels.
  • The method incorporates techniques to handle ill-conditioning during reconstruction.

Main Results:

  • WLR successfully compensated for structured pruning without introducing extra parameters.
  • The method demonstrated improved numerical stability and efficiency in pruned LLMs.
  • Evaluations on LLaMA and OPT models showed performance exceeding state-of-the-art methods.

Conclusions:

  • Well-conditioned Linear Reconstruction (WLR) offers an effective parameter-free compensation strategy for structured LLM pruning.
  • WLR enhances model stability and performance, making it suitable for efficient deployment.
  • The proposed method represents a significant advancement in optimizing large language models.