DeepSeek เปิดตัวโมเดลคณิตศาสตร์ DeepSeek-Math-V2 โอเพนซอร์ส คว้าเหรียญทอง IMO 2025

#ShoperGamer #DeepSeek #Ai #ModelAi #OpenSource #DeepSeekMathV2 #Feed

DeepSeekMath‑V2 is an open‑source beast that writes and checks its own math proofs, using an internal critique loop and low test‑time compute. Could it be the pop‑off point for the US AI bubble? Dive in to see how it stacks against OpenAI and DeepMind. #DeepSeekMathV2 #OpenSourceAI #MathProofs

🔗

118 zu 90 Punkte gegen Menschen:

- DeepSeekMath-V2 schlägt Top-Studenten im Putnam-Test
- Selbstkontrolle durch Meta-Verifier Architektur
- Komplett Open Source unter Apache 2.0

Ist Mathe als menschliche Fähigkeit damit offiziell obsolet? #DeepSeekMathV2

DeepSeekMath-V2 ile matematiksel akıl yürütmede yeni bir çağ! Kendi kendini doğrulayan yapısıyla teorem ispatlamada devrim. Bilim ve teknoloji için önemli bir adım. Detayları inceleyin!\n\n🚩 #DeepSeekMathV2 #YapayZeka #Matematik #Bilim

winbuzzer.com/2025/11/27/d...

DeepSeekMath-V2 Matches OpenAI and Google with IMO Gold Medal Win

#AI #DeepSeek #OpenSourceAI #GenAI #MathAI #ChinaAI #AIReasoning #IMO2025 #DeepSeekMathV2

¡DeepSeekMath-V2 lo logra! La IA alcanza puntajes de Oro en IMO 2025 y CMO 2024. Su clave: Razonamiento matemático auto-verificable y dominio en prueba de teoremas. ¿El futuro de la IA científica? youtu.be/XXiFWieCFao #DeepSeekMathV2 #LLMs #IA #Matematicas #TheoremProving

Wow! Nous Research's Nomos 1 just took 2nd place on the Putnam, right behind DeepSeekMath‑V2. Can open‑source AI really challenge Google Gemini and OpenAI? Dive into the showdown and see the numbers. #Nomos1 #DeepSeekMathV2 #Putnam

🔗 aidailypost.com/news/nous-re...

DeepSeek Math V2 just dropped a dual‑component setup with a generator & verifier, plus a two‑stage reward‑model training. Open‑source proof generation just got smarter. Dive into the details! #DeepSeekMathV2 #DualComponent #TwoStageTraining

🔗 aidailypost.com/news/deepsee...

DeepSeekMath‑V2 just snagged gold at IMO 2025 and crushed the China Math Olympiad—outperforming Putnam 2024 champs and AIME pros. This open‑source AI is rewriting theorem‑proving playbooks. Curious? Dive into the details! #DeepSeekMathV2 #IMO2025 #AITheoremProving

🔗 aidailypost.com/news/deepsee...

DeepSeekMathV2: a IA chinesa que redefiniu os limites do raciocínio matemático:

Chinese startup #Deepseek’s new #DeepseekMathV2 model achieved #goldmedal-level results at the International Mathematical Olympiad (#IMO) 2025 and the Chinese #CMO 2024. The model uses a #multistageprocess with a “verifier” and “meta-verifier” to evaluate and refine its own solutions, achieving…