← Back to live feed1 story
Coinbase fine-tuned the open weight Qwen3.5-9B model on proprietary data to identify fraudulent activity, outperforming generalist frontier models in speed and cost. The company has implemented an internal harness allowing employees to train their own models for tasks such as payments risk review.
Sign in to suggest edits
Key sources
- SOURCE@brian_armstrong“fine tune an open weight model (Qwen3.5-9B) on proprietary internal data at @coinbase, and it out performed frontier models, at much lower cost and speed, for a specialized task (fraud prevention in this case)”x.com
- SOURCEmarketbrief.now