📄 文章列表

3985 篇文章
cn_ai

同一个模型从 30 分到 95 分：拉开差距的不是模型，是Harness

✅ completed ⭐ ☆☆☆☆☆ 2026-11-04
cn_ai

Harbor 新增 Docker Sandboxes 支持，可在隔离的 MicroVM 中运行可重现的 AI 代理评估

✅ completed ⭐ ☆☆☆☆☆ 2026-11-04
cn_ai

Anthropic 为重要开源项目新增可选安全扫描服务

✅ completed ⭐ ☆☆☆☆☆ 2026-11-04
AlphaSignal

Snorkel AI's RIFT Exposes Hidden Flaws in 83% of AI Benchmark Rubrics

✅ completed ⭐ ☆☆☆☆☆ 2026-11-02
AlphaSignal

Ai2's StateBench Catches Frontier Models Failing at 13% on Long State Tracking

✅ completed ⭐ ☆☆☆☆☆ 2026-11-03
AlphaSignal

Anthropic Kills Repetitive Login Screens for Claude's MCP Connectors

✅ completed ⭐ ☆☆☆☆☆ 2026-11-03
AlphaSignal

Anthropic's FlashProve Cryptographically Proves Which AI Model Answered You

✅ completed ⭐ ☆☆☆☆☆ 2026-11-03
AlphaSignal

Artificial Analysis Opens CritPt v2 API so Developers Can Grade Physics AI

✅ completed ⭐ ☆☆☆☆☆ 2026-11-03
AlphaSignal

Boston Dynamics Rebuilds Stretch Gen 2 to Unload 1,000 Boxes per Hour

✅ completed ⭐ ☆☆☆☆☆ 2026-11-03
AlphaSignal

Google Runs Gemma 4 on a TPU in Orbit for the First Time

✅ completed ⭐ ☆☆☆☆☆ 2026-11-03
AlphaSignal

Krea Releases Krea 3, Beating GPT Image 2.5 at 11x Lower Latency

✅ completed ⭐ ☆☆☆☆☆ 2026-11-03
AlphaSignal

Sarvam's Prism Turns a 105B Open Model Into a 44ms Decision Engine

✅ completed ⭐ ☆☆☆☆☆ 2026-11-03
AlphaSignal

Topaz Labs' Wonder 4 Fixes and Enlarges Photos in One Smart Pass

✅ completed ⭐ ☆☆☆☆☆ 2026-11-03
AlphaSignal

Artificial Analysis Rebuilds Its AI Capability Indexes, Crowning Claude Opus 5.5

✅ completed ⭐ ☆☆☆☆☆ 2026-11-04
VentureBeat

Open-weight AI models are getting cheaper and better. Here's how to use them without taking on the cost of running them

✅ completed ⭐ ☆☆☆☆☆ 2026-10-31
cn_ai

从英伟达到 Pika，硅谷巨头为何“反向借力”这个国产模型？

✅ completed ⭐ ☆☆☆☆☆ 2026-11-03
cn_ai

AI 改变的是数据接口，而不是查询引擎

✅ completed ⭐ ☆☆☆☆☆ 2026-11-03
cn_ai

A社紧急修改招股书！RSI我们快要控制不住了

✅ completed ⭐ ☆☆☆☆☆ 2026-11-03
AlphaSignal

Anthropic Pays $500K to Lure Professors and Founders Into AI Safety

✅ completed ⭐ ☆☆☆☆☆ 2026-11-02
AlphaSignal

Anthropic's Message Threads Cut Agent Request Size from 800 KB to 1.8 KB

✅ completed ⭐ ☆☆☆☆☆ 2026-11-02