Alibaba/Qwen

Latest AI news, models and releases from Alibaba/Qwen. ['CodeQwen1.5', 'Cotype Light 3', 'Cotype Pro 3', 'Fun-Realtime-TTS', 'Hanguang', 'Hanguang 800', 'HappyOyster 1.0', 'ICN Switch 1.0', 'not specified', 'Panjiu', 'Qianwen', 'Qoder Security', 'QVQ-72B-Preview', 'QVQ-Max', 'Qwen', 'Qwen 2', 'Qwen2.5', 'Qwen2.5-0.5B', 'Qwen2.5-14B', 'Qwen2.5-14B-Instruct-1M', 'Qwen2.5-1.5B', 'Qwen2.5-32B', 'Qwen2.5-3B', 'Qwen-2.5 72B', 'Qwen2.5-72B', 'Qwen2.5-7B', 'Qwen2.5-7B-Instruct-1M', 'Qwen2.5-Coder', 'Qwen2.5-Coder-0.5B', 'Qwen2.5-Coder-0.5B-Instruct', 'Qwen2.5-Coder-14B', 'Qwen2.5-Coder-1.5B', 'Qwen2.5-Coder-1.5B-Instruct', 'Qwen2.5-Coder-32B', 'Qwen2.5-Coder-32B-Instruct', 'Qwen2.5-Coder-3B', 'Qwen2.5-Coder-3B-Instruct', 'Qwen2.5-Coder-7B', 'Qwen2.5-Coder-7B-Instruct', 'Qwen2.5-Coder-Instruct']

Official site ↗

AI Safety 🇺🇸

AI models prone to 'overthinking' creates security vulnerability

Researchers from Zhejiang University and Alibaba have demonstrated a method to deliberately induce 'overthinking' in reasoning AI models by subjecting them to logically inconsistent prompts. This evolutionary prompt attack can cause outputs up to 26 times longer, effectively acting as a denial-of-service attack on commercial AI services.

DeepSeek

Alibaba/Qwen

OpenAI

Google/DeepMind

IEEE Spectrum AI27.07 · 04:05

Research 🇷🇺

QuantCode-Bench: a new benchmark for evaluating LLMs' ability to generate executable algorithmic trading strategies

Researchers developed QuantCode-Bench, a benchmark to assess how well large language models (LLMs) can generate executable algorithmic trading strategies from textual specifications. The benchmark evaluates not only code correctness but also semantic alignment with the original trading idea through four stages: compilation, backtest, trade generation, and judge verification.

OpenAI

Anthropic

Google/DeepMind

xAI

DeepSeek

Alibaba/Qwen

Moonshot AI

Habr — хаб NLP27.07 · 03:03

Show more ↓