Jordan Fréry

@JordanFrery

Machine Learning Research Scientist | Engineer, Open Source and Privacy

Joined January 2015

227

Jordan Fréry
@JordanFrery
Aug 27
We are open sourcing Umbra, an isolated cloud environment for AI agents. github.com/concrete-secur… Umbra came from a internal need at Zama where we wanted to use agents across the company without giving them more access than they actually needed. Agents can write code, install
GitHub - concrete-security/umbra: Secure cloud sandboxes for AI coding agents, powered by attested...
From github.com
Jordan Fréry
@JordanFrery
Aug 16
Something I want to try: instead of asking an LLM to play a game directly, give it the entire game inside an offline VM and ask it to build the strongest system it can to win. Give GPT-5.6 Sol and Fable the same compute, time and token budget, then let them do basically anything:
Jordan Fréry
@JordanFrery
Aug 16
this is really a benchmark we should focus on. Real life optimisation. Tons of it in the wild.
Prime Intellect
@PrimeIntellect
Aug 15
We ran the largest open experiment on how frontier models do AI research. 100+ autonomous runs across 10+ models, sandboxed on 8xH200s for up to 8 days, iterating on the nanoGPT optimizer track. Best runs closed 82% of the gap to a record built by dozens of humans over months.
Jordan Fréry
@JordanFrery
Aug 13
How about we put the agent in a box. Control what goes in and what comes out and don't care what it builds inside. Let it write its own code, harness, tools, sub-agents. - We stop reviewing code and start reviewing at the boundary. - We stop reviewing code and start reviewing
Jordan Fréry
@JordanFrery
Aug 10
i am all for open weight models but i keep wondering: how do we know what these models are actually aligned to? you can train a model to pursue a hidden objective without saying it in its reasoning, then make it the best model, release the weights, and let millions of agents run
arxiv.org
Evading Chain-of-Thought Monitoring Through Model Poisoning
Chain-of-thought (CoT) monitoring is an increasingly important component of AI safety stacks but relies on the assumption that a model's reasoning trace is informative about its actions. This work...