Passer au contenu Passer à l'index du site

Gagnez jusqu'à 2 000 $US lorsque vous achetez 50 $US de cryptos.¹

Design Improvements

Jul 2, 2013 • Lecture de 1 min

Our transaction show page got a small face lift today. Design is something we’ve historically under invested in at Coinbase, but hopefully that should be changing soon. Stay tuned for some coming updates!

Articles récents

Singapore Sees 43% Drop in Crypto Scam Losses and Partnerships Like Ours Are Helping Drive the Change

Coinbase

Oct 13, 2026•Lecture de 2 min

TL;DR: Singapore’s crypto scam losses fell 43% year over year, from S$114.9 million to S$65.5 million, as a result of stronger enforcement, public education, and partnerships between the Anti-Scam Centre and industry leaders like Coinbase. Through four joint operations, Coinbase’s Global Intelligence team helped identify 720+ victims and prevent more than S$18.9 million in potential losses, while a new anti-scam awareness campaign reached Singapore retail customers.

International

Part 3/3 Owning Intelligence Series: How Coinbase Built, Evaluated, and Post-Trained a Fraud Agent

Yao Ma, Xuwei Tan, Akshit Trehan, Aman Choudhary, Anyscale Team: Eric Tang, Sumanth Hegde, Seiji Eicher, Jess Kong

Oct 8, 2026

TL;DR: Using SkyRL on Anyscale, we post-trained Qwen3.5-9B for Onramp fraud detection and served it with Ray Serve LLM and vLLM. On our benchmark, it exceeded Opus 4.5 across all four detection metrics, including 9.6% higher F1. In online inference, we reduced end to end latency by 55%.

Engineering

Part 2/3 Owning Intelligence Series: How Coinbase Built, Evaluated, and Post-Trained a Fraud Agent

Yao Ma, Xuwei Tan

Oct 7, 2026

TL;DR: Newer LLMs are not necessarily better at domain-specific tasks. On our fraud benchmark, newer Opus, Sonnet, and GPT versions all had lower recall, F1, and dollar-weighted recall—even when precision improved. The lesson: choose models using high-quality, domain-specific benchmarks, with reproducible evaluation and ideally deterministic scoring against verified outcomes. These findings also motivated our next step: post-training open-weight models for greater control over quality, cost, and deployment.

Engineering

Informations : Le trading de produits dérivés via la plateforme Coinbase Advanced est disponible pour les client(e)s éligibles de l'EEE (Espace économique européen) par l'intermédiaire de Coinbase Financial Services Europe Ltd. (Licence CySEC 374/19). Afin de pouvoir accéder aux produits dérivés, les client(e)s devront réussir une évaluation de niveau visant à déterminer leur éligibilité à ce type de produit (et s'il est adapté à leurs besoins).

Design Improvements

Articles récents

Working Together on Bitcoin’s Post-Quantum Readiness

Consumer Protection Tuesday: Following the money. Stopping the scammers.

Singapore Sees 43% Drop in Crypto Scam Losses and Partnerships Like Ours Are Helping Drive the Change

Part 3/3 Owning Intelligence Series: How Coinbase Built, Evaluated, and Post-Trained a Fraud Agent

Coinbase and Samsung Bring USDC to Samsung Wallet

Part 2/3 Owning Intelligence Series: How Coinbase Built, Evaluated, and Post-Trained a Fraud Agent