📄 文章列表

35 篇文章
cn_ai

谷歌推出 EnvHarness，支持自进化智能体动态环境学习

✅ completed ⭐ ☆☆☆☆☆ 2026-10-30
cn_ai

不再逐个生成 Token：扩散模型迈向 35B，开始挑战 Agent 编程

✅ completed ⭐ ☆☆☆☆☆ 2026-10-29
cn_ai

刚刚，扩散模型杀进Agent主场！350亿参数，速度比Jev快一倍

✅ completed ⭐ ☆☆☆☆☆ 2026-10-29
TDS

Is the #1 AI Model Really #1?

✅ completed ⭐ ☆☆☆☆☆ 2026-10-29
VentureBeat

A coding agent’s harness can triple its cost without making it more accurate, study finds

✅ completed ⭐ ☆☆☆☆☆ 2026-10-23
VentureBeat

New MIT and Sakana AI framework uses an LLM judge to cut evaluation costs for self-improving coding agents

✅ completed ⭐ ☆☆☆☆☆ 2026-10-02
VentureBeat

Google’s open source EnvHarness lets AI agents train against environments that evolve with them

✅ completed ⭐ ☆☆☆☆☆ 2026-09-21
VentureBeat

Salesforce researchers took an AI agent from finishing 43.5% of browser tasks to 93% without touching the model

✅ completed ⭐ ☆☆☆☆☆ 2026-09-16
AlphaSignal

xAI's Grok Build Now Lets Coding Agents Search 500B URLs Instantly | AlphaSignal

✅ completed ⭐ ☆☆☆☆☆ 2026-07-24
AlphaSignal

GitHub Copilot Matches Claude Code While Burning Fewer Tokens | AlphaSignal

✅ completed ⭐ ☆☆☆☆☆ 2026-07-12
AlphaSignal

OpenAI Retracts SWE-Bench Pro After Finding 30% of Tasks Broken | AlphaSignal

✅ completed ⭐ ☆☆☆☆☆ 2026-07-08
TDS

How to Choose Between Small and Frontier Models

✅ completed ⭐ ☆☆☆☆☆ 2026-06-29
AlphaSignal

Agentic Tests Beyond the Bug Hunt: DeepSWE, ProgramBench, and ITBench-AA push agents harder than SWE-bench

✅ completed ⭐ ★★★★★ 2026-06-19
AlphaSignal

Testing Mythos and Fable, Moving Beyond SWE-bench, Nvidia's Open Contender

✅ completed ⭐ ★★★★★ 2026-06-19
Rundown.AI

OpenAI's best model just got better

✅ completed ⭐ ★★★★★★ 2024-08-21
Rundown.AI

Google's advanced AI voice is here!

✅ completed ⭐ ★★★★★★ 2024-08-14
AlphaSignal

Prime Intellect Offers Free RL Fine-Tuning Runs on Poolside's Laguna XS.2 | AlphaSignal

✅ completed ⭐ ★★★★★ 2026-06-26
AlphaSignal

Nous Research Drops Step 3.7 Flash Free Inside the World's Fastest-Growing Agent | AlphaSignal

✅ completed ⭐ ★★★★★ 2026-06-26
AlphaSignal

Anthropic's Claude Fable 5 Tops the Hardest Coding Benchmark With 70% Score | AlphaSignal

✅ completed ⭐ ★★★★★ 2026-06-26
AlphaSignal

Anthropic's Claude Fable 5 Now Powers Browser Use Terminal's Autonomous Web Agent | AlphaSignal

✅ completed ⭐ ★★★★★ 2026-06-26