#gpt-2

[ follow ]
fromHackernoon
1 year ago

Empirical Results: GPT-2 Analysis of Transformer Memorization & Loss | HackerNoon

These experiments with GPT-2 medium on OpenWebText validate the radius hypothesis from our theoretical framework, measuring activation distances in the last layer for next-token prediction.
Roam Research
fromHackernoon
1 year ago

GPT-2 Study Shows How Language Models Can Amplify Political Bias | HackerNoon

This study highlights the critical issue of bias amplification in large language models (LLMs), demonstrating its impact predominantly through the lens of political bias in U.S. media.
Artificial intelligence
fromHackernoon
7 months ago

How Tokenizer Choices Shape Hidden Risks in Popular Language Models | HackerNoon

Tokenization methods can reveal under-trained tokens that impact model performance.
Identifying and addressing under-trained tokens is crucial for improving language model accuracy.
Privacy professionals
fromTechCrunch
4 months ago

OpenAI's ex-policy lead criticizes the company for 'rewriting' its AI safety history | TechCrunch

Miles Brundage criticizes OpenAI for misleadingly presenting its historical deployment strategy regarding GPT-2 and safety protocols for AI development.
[ Load more ]