AI ethics evidence index Search
Source-linked daily record
Evidence record 3331 · automatically gathered

What Makes a Strong Model? A Unified Spectral Analysis of Knowledge Transfer over High-dimensional Linear Regression

Teacher-Student Knowledge Transfer (KT) is ubiquitous in modern machine learning, ranging from classical model compression via Knowledge Distillation (KD) to the emergent phenomenon of Weak-to-Strong (W2S) generalization. While existing studies offer isolated insights, a unified theoretical framework explaining the efficacy of KT across these disparate regimes remains lacking. In this work, we establish a unified spectral analysis of SGD dynamics in high-dimensional linear regression, elucidatin

Record details

Published: 31 May 2026
Source: arXiv
Category: Research
Topics: Bias & fairness · Children & education
Retrieved: 14 July 2026

source-onlyevidence status

These records share source-supplied organisations, an exact publisher byline, automatic topics or regions. The reason is shown on every link; related does not mean supporting, agreeing with or verifying this record.

Shared topic: Bias & fairness · Shared topic: Children & education

TriEval: A Resource-Efficient Pipeline for LLM Bias, Toxicity, and Truthfulness Assessment

arXiv · 2 June 2026

Shared topic: Bias & fairness · Shared topic: Children & education

Two-Phase Simulated Annealing for Equitable Team Formation: Eliminating Complaints in Large Engineering Cohorts

arXiv · 5 June 2026

Shared topic: Bias & fairness · Shared topic: Children & education

Impacts of Histories and Models on LLM Grading: A Study in Advanced Software Engineering Courses

arXiv · 7 June 2026

Shared topic: Bias & fairness · Shared topic: Children & education

Using AI in engineering education: a balancing act, driven by clear purpose

arXiv · 15 June 2026

Shared topic: Bias & fairness · Shared topic: Children & education

Context Matters: Auditing Gender Bias in T2I Generation through Risk-Tiered Use-Case Profiles

arXiv · 13 May 2026

Shared topic: Bias & fairness · Shared topic: Children & education

Preparing Students for AI-Powered Materials Discovery: A Workflow-Aligned Framework for AI Literacy, Equity, and Scientific Judgment

arXiv · 10 May 2026

How to cite this record

ethics.ai (31 May 2026), “What Makes a Strong Model? A Unified Spectral Analysis of Knowledge Transfer over High-dimensional Linear Regression,” evidence record 3331, https://ethics.ai/record/3331 (originally published by arXiv).

JSON

Use and limitations

This page is a stable index and citation surface for a source record. ethics.ai did not author the underlying report and has not independently verified every claim. Automatic topics may be imperfect. For consequential use, quote and cite the original publisher.