Open Models, Closed Gaps: How Fine-Tuning Impacts AI Model Toxicity
9 Jun 2025
This study explores how fine-tuning impacts toxicity in open-source language models, backed by reproducible experiments and open-access code.
Why AI Models Get More Toxic After Community Fine-Tuning
9 Jun 2025
Fine-tuning AI models can unexpectedly increase toxicity—even with non-adversarial data—raising concerns for developers and policymakers alike.
Fine-Tuning Can Accidentally Make AI More Toxic, Study Finds
9 Jun 2025
Fine-tuning can unintentionally undo AI safety work, increasing toxicity—even without harmful data. Safety must be re-evaluated after each tweak.
Multilingual AI Fine-Tuning Shows Mixed Results on Toxicity
9 Jun 2025
AI models fine-tuned by the community show unpredictable toxicity levels, with results varying across languages and tuning approaches.
Can AI Be Taught to Be Less Toxic? New Findings Say Yes (But...)
9 Jun 2025
Instruction tuning reduces AI model toxicity, but additional tuning with the Dolly dataset may unintentionally increase harmful outputs in some models.
Community-Tuned AI Models Are Popular—But Are They Safe?
9 Jun 2025
How do fine-tuning and community variants affect AI toxicity? A study of 28 small language models reveals surprising shifts in toxic output rates.
The Dark Side of AI Fine-Tuning
9 Jun 2025
Fine-tuning boosts AI performance—but at what cost? This article explores how tuning can backfire, increasing model toxicity and undermining safety.
How Fine-Tuning Open AI Models Can Reintroduce Toxicity
9 Jun 2025
Small fine-tuning changes in open AI models like Llama and Gemma can undo safety measures—leading to unpredictable and toxic outputs.
Tracking Economic Sentiment: Advanced Time Series Models for Survey Responses
15 May 2025
GAR(1) models reveal trends in consumer inflation beliefs, showing how metric-space time series illuminate economic perceptions over time.