📄 文章列表
89 篇文章
AlphaSignal
Agentic Tests Beyond the Bug Hunt: DeepSWE, ProgramBench, and ITBench-AA push agents harder than SWE-bench
AlphaSignal
Top Agentic Performance, Low Cost: GLM-5.2, designed for coding and long-running agentic jobs, now the top open model
AlphaSignal
Claude Opus 4.8 advertises its uncertainty: Dynamic workflows manage Anthropic subagents
AlphaSignal
Claude Fable 5, or Mythos for the masses: Apple rethinks mixture-of-experts to save local memory
AlphaSignal
DiffusionGemma breaks 1,000 tokens/second: Anthropic closes Fable and Mythos after US intervention
AlphaSignal
Zhipu’s GLM-5.2 is the new top open model: SpaceX buys coding favorite Cursor
AlphaSignal
U.S. chatbot use passes 50 percent: AA-Briefcase benchmark measures knowledge work
AlphaSignal
Inside Jalapeño, OpenAI’s first inference chip: Model synthesis approaches frontier performance
AlphaSignal
Mythos Begets Fable, Cursor's Composer 2.5, Agents Building Agents
AlphaSignal
Testing Mythos and Fable, Moving Beyond SWE-bench, Nvidia's Open Contender
AlphaSignal
A New Generation Studies AI, Apple's Recipe for On-Device Models, GLM5.2 Tackles Open-Ended Problems
AlphaSignal
Cursor Catches Composer 2.5 and Claude Opus Gaming Coding Benchmarks | AlphaSignal
AlphaSignal
Cognition Ships Claude Fable 5 Across All Devin Tiers, Doubling Top Benchmark Scores | AlphaSignal
AlphaSignal
Artificial Analysis' AA-Briefcase Benchmark Exposes What AI Models Can't Actually Do | AlphaSignal
AlphaSignal
Anthropic's Claude Opus 4.8 Beats Its Predecessor but Trails GPT-5.5 by 12 Points | AlphaSignal
AlphaSignal
Anthropic's Claude Fable 5 Tops AI Charts but Silently Routes Sensitive Tasks Away | AlphaSignal
AlphaSignal
Anthropic's Claude Fable 5 Beats the Hardest Coding Benchmark in Under 2 Hours | AlphaSignal
AlphaSignal
Anthropic's Claude Fable 5 Beats Every Rival but Blocks Itself on Dangerous Requests | AlphaSignal
AlphaSignal
Warp Adds Claude Opus 4.8 With 4x Better Code Error Detection | AlphaSignal
AlphaSignal