Top Tweets for #exploitbench

7 days ago

The "critical cybersecurity threshold" that @OpenAI flagged for #Astra means the model is capable enough to find and exploit real vulnerabilities. That's why #ExploitBench exists. That's why #Mythos 5.1 is gated. The frontier labs are running their own red teams on their own models before release. That's not a feature. It's the new normal.

TechPP @techpp

27 days ago

GPT-6 Astra (xhigh) leads the ExploitBench ranking with 100% 1. GPT-6 Astra (xhigh) · 100% 2. GPT-5.6 Sol (max) · 78.5% #ExploitBench #OpenAI #GPT6Astraxhigh #GPT56Solmax

techpp's tweet photo. GPT-6 Astra (xhigh) leads the ExploitBench ranking with 100%

1. GPT-6 Astra (xhigh) · 100%
2. GPT-5.6 Sol (max) · 78.5%

#ExploitBench #OpenAI #GPT6Astraxhigh #GPT56Solmax https://t.co/FALnIP9H0h

235

Top Tweets for #exploitbench

Last Seen Hashtags on Sotwe

Trends for you

Most Popular Users