[tag]

#llm-evaluation

2 stories

[research]Aug 31, 2026By ByteBulletin Editor

Rasch Models Bring a Statistical Lens to LLM Evaluation

A new framework applies Rasch measurement theory to benchmark large language models, offering a more rigorous way to compare model proficiency and item difficulty.

[research]Aug 30, 2026By ByteBulletin Editor

WebGrader: An Automated Tool for Evaluating LLM-Generated Web Code

A new framework uses automated evaluation to grade web code generated by large language models, moving beyond manual review.