KDD Cup champions win using only DeepSeek web chat interface
jiqizhixin · x · 2026-08-19
At the TAAC x KDD Cup 2026, the industrial track champions utilized only a single DeepSeek web chat session—no APIs, GPT, or Claude—to design a unified Recommendation Block and optimize AUC. The competition leveraged anonymized data from Tencent's real-world business serving billions of users, exploring the Scaling Law of recommendation systems. Meta's Rui Li noted that unlike LLMs, recommendation systems lack established benchmarks, making this real-world problem exposure significant.
Related event: KDD Cup Winners Used Only DeepSeek Chat to Take the Crown(4 posts)→
More from Research
- Peking U & Kling Team Release RefCaptioner, Tackling Video Understanding Hallucinations — jiqizhixin · 2026-08-19
- 'Scaling laws are not laws of nature': better data and architectures can still bend the curves — bookwormengr · 2026-08-19
- PixRestore: Unified Image Restoration via Pixel Diffusion Transformer — Lingchen Sun · 2026-08-19
- Cross-Model Memory Transfer via Target-Side Reader Adaptation — OLAResearchX · 2026-08-19
- DeepSeek Web Chat Wins Industrial Track in TAAC × KDD Cup — jiqizhixin · 2026-08-19
- Next-Gen Agent Frameworks: Auto-Integrated Coded Extensions — _philschmid · 2026-08-19