Marc Fischer
@marc_r_fischer
Followers
427
Following
4K
Statuses
4K
Co-Founder of @InvariantLabsAI, PhD student at ETH Zurich. I care about security and reliability of AI systems. @[email protected]
Zurich, Switzerland
Joined March 2010
RT @InvariantLabsAI: If you want to play with some agent tooling over the holidays, check out our Invariant Winter challenge. Can you fix…
0
1
0
RT @InvariantLabsAI: 🎄 Ho ho ho 🎅 An early holiday celebration from Invariant Labs! We are launching (and open-sourcing!) Explorer, a visua…
0
3
0
RT @InvariantLabsAI: This week, AgentDojo is presented at #NeurIPS2024. AgentDojo helps benchmark agents, leading to secure and reliable AI…
0
3
0
Check out our new work from @InvariantLabsAI. Many agent builders in academia and industry told me that they had never looked at the submissions of standard agent benchmarks because it is cumbersome. We fixed that.
🤖 Something new and exciting: We have created a public registry of AI agent benchmarks! 🌐 Check it out at
0
0
4
@Cryptomaeher @zornsllama Had the same procedure done by a dentist that shares your last name 😬
1
0
1
RT @InvariantLabsAI: This week, Lavinia Petrache joined us as our first full-time employee and Head of Product. 🧚🏻♀️ Before joining Invar…
0
1
0
RT @InvariantLabsAI: 🚀 Invariant Labs is now an official @ETH_en Spin-Off! We build on our research on AI Security conducted at ETH over th…
0
1
0
Think you know AI security? Try our challenge!
🚀 Excited to announce the first Capture The Flag (CTF) challenge for security of AI agents with a $1000 prize pool! What happens if a customer accidentally posts a password into a feedback form, which is then analyzed by an AI agent and posted into a private Discord channel?
0
0
1
RT @rajko_rad: Fourth down and a wrap! M̴a̴t̴h̴ ̴x̴ ̴C̴o̴d̴e̴ ̴x̴ ̴A̴g̴e̴n̴t̴s̴ Cc @amanrsanger (@cursor_ai ) @AlbertQJiang (@MistralAI )…
0
5
0
RT @InvariantLabsAI: 👉 Tomorrow Invariant Labs researchers are presenting our paper "AI Agents with Formal Security Guarantees" at the @icm…
0
3
0
On Friday, at the @NG_AI_Safety workshop, we will show work from @InvariantLabsAI on making AI Agents secure! More details will follow! 👨💻 Work with @mbalunovic, @lbeurerkellner, @mvechev.
0
0
3
RT @edoardo_debe: Does the instruction hierarchy introduced with GPT-4o mini work? We ran AgentDojo on it, and it looks like it does! GPT-…
0
5
0
RT @InvariantLabsAI: We are excited to welcome the newest addition to the Invariant Labs team – Dragoș Albăstroiu @adragos_, who is joining…
0
3
0
Next week I'm gonna be at #ICML2024 (@icmlconf) in Vienna! Reach out if you want to talk about AI Agents, AI Agent Security, Constrained Decoding for LLMs, or @InvariantLabsAI. Also, hit me up if you want so recommendations for Vienna (coffee, drinks, restaurants, culture...)
0
0
12