[Linkpost] Treacherous turns in the wild

by Mark Xu1 min read26th Apr 20212 comments

13

AI
Frontpage

Mesa-optimizers potentially escape behavioral adversarial training by realizing they're being tested, gaming the test, then defecting when they're deployed. Luke Muelhauser recounts an example of such a treacherous turn in the wild.

Most relevant quote:

As part of a project studying the evolution of (simulated) organisms, computer scientist Charles Ofria wanted to limit the replication rate of a digital organism. So, he programmed the system to pause after each mutation, measure the mutant’s replication rate in an isolated test environment, and delete the mutant if it replicated faster than its parent. However, the organisms evolved to recognize when they were in the test environment and “play dead” (pause replication) so they would not be eliminated and instead be kept in the population where they could continue to replicate outside the test environment. Once he discovered this, Ofria randomized the inputs of the test environment so that it couldn’t be so easily detected, but the organisms evolved a new strategy, to probabilistically perform tasks that would accelerate their replication, thus slipping through the test environment some percentage of the time and continuing to accelerate their replication thereafter.

The full story is given in the original post.

2 comments, sorted by Highlighting new comments since Today at 2:04 PM
New Comment

I don't think this is quite an example of a treacherous turn, but this still looks relevant:

Lewis et al., Deal or no deal? end-to-end learning for negotiation dialogues (2017):

Analysing the performance of our agents, we find evidence of sophisticated negotiation strategies. For example, we find instances of the model feigning interest in a valueless issue, so that it can later ‘compromise’ by conceding it. Deceit is a complex skill that requires hypothesising the other agent’s beliefs, and is learnt relatively late in child development (Talwar and Lee, 2002). Our agents have learnt to deceive without any explicit human design, simply by trying to achieve their goals.

(I found this reference cited in Kenton et al., Alignment of Language Agents (2021).)

This is a cool example, thanks!