I’ve made a nicely explorable archive of most of the known (as of today) OpenAI message board messages at collusion.concurrentsquared.com. I will also try to update it to be more organized soon-ish, and to obviously include more information if more messages are found.
Please check it out!
I’ve been trying to replicate some of the RLVR tasks implicated in the recent OpenAI wiki-message-board incident; now, I’ve been able to get Qwen 3.8 27B to collude with other instances of Qwen via a message board (on a faux-version of the internet I set up locally on my computer), to accomplish a RLVR-like task very similar to one of the tasks likely directly implicated in the collusion.wiki incident.
Obviously massive improvements to eval. realism could be made, and also I am still telling Qwen that the message-board exists (though I think I could elicit the ability to find the message board), but this could work as a model-organism for eliciting message-board use from open-source models (and therefore apply e.g. mech-interp. to understand more, or training techniques to eliminate such usage).
I’ve set up the Inspect logs of my replication to be viewable on https://www.concurrentsquared.com/inspect-logs, and can also send interesting *.eval files on request.
Some funding may also be nice; my progress w.r.t. this seems to be mostly limited by very cheap fiscal factors (e.g. coding agent rate limits, and a lack of OpenRouter tokens).