by Tom Everitt 7 days ago | link | parent Hi Vadim! So basically the advisor will be increasingly careful as the cost of falling into the trap goes to infinity? Makes sense I guess. What is the incentive for the agent not to always let the advisor choose? Is there always some probability that the advisor saves them from infinite loss, or only in certain situations that can be detected by the agent?

 by Vadim Kosoy 6 days ago | link If the agent always delegates to the advisor, it loses a large fraction of the value. Returning again to the simple example above, the advisor on its own is only guaranteed to get expected utility $$1/2 + \omega(t^{-1/3})$$ (because it often takes the suboptimal action 1). On the other hand, for any prior over a countable set of environments that includes this one, the corresponding DIRL agent gets expected utility $$1 - o(1)$$ on this environment (because it will learn to only take action 2). You can also add an external penalty for each delegation, adjusting the proof is straightforward. So, the agent has to exercise judgement about whether to delegate, using its prior + past observations. For example, the policy I construct in Lemma A delegates iff there is no action whose expected loss (according to current beliefs) is less than $$\beta(t)^{-1}t^{-1/3}$$. reply
 by Tom Everitt 6 days ago | link So this requires the agent’s prior to incorporate information about which states are potentially risky? Because if there is always some probability of there being a risky action (with infinitely negative value), then regardless how small the probability is and how large the penalty is for asking, the agent will always be better off asking. (Did you see Owain Evans recent paper about trying to teach the agent to detect risky states.) reply
 by Vadim Kosoy 5 days ago | link The only assumptions about the prior are that it is supported on a countable set of hypotheses, and that in each hypothesis the advisor is $$\beta$$-rational (for some fixed $$\beta(t)=\omega(t^{2/3})$$). There is no such thing as infinitely negative value in this framework. The utility function is bounded because of the geometric time discount (and because the momentary rewards are assumed to be bounded), and in fact I normalize it to lie in $$[0,1]$$ (see the equation defining $$\mathrm{U}$$ in the beginning of the Results section). Falling into a trap is an event associated with $$\Omega(1)$$ loss (i.e. loss that remains constant as $$t$$ goes to $$\infty$$). Therefore, we can risk such an event, as long as the probability is $$o(1)$$ (i.e. goes to $$0$$ as $$t$$ goes to $$\infty$$). This means that as $$t$$ grows, the agent will spend more rounds delegating to the advisor, but for any given $$t$$, it will not delegate on most rounds (even on most of the important rounds, i.e. during the first $$O(t)$$-length “horizon”). In fact, you can see in the proof of Lemma A, that the policy I construct delegates on $$O(t^{2/3})$$ rounds. As a simple example, consider again the toy environment from before. Consider also the environments you get from it by applying a permutation to the set of actions $$\mathcal{A}$$. Thus, you get a hypothesis class of 6 environments. Then, the corresponding DIRL agent will spend $$O(t^{2/3})$$ rounds delegating, observe which action is chosen by the advisor most frequently, and perform this action forevermore. (The phenomenon that all delegations happen in the beginning is specific to this toy example, because it only has 1 non-trap state.) If you mean this paper, I saw it? reply

### NEW DISCUSSION POSTS

I have stopped working on
 by Scott Garrabrant on Cooperative Oracles: Introduction | 0 likes

 by Vadim Kosoy on Delegative Inverse Reinforcement Learning | 0 likes

So this requires the agent's
 by Tom Everitt on Delegative Inverse Reinforcement Learning | 0 likes

If the agent always delegates
 by Vadim Kosoy on Delegative Inverse Reinforcement Learning | 0 likes

 by Tom Everitt on Delegative Inverse Reinforcement Learning | 0 likes

Hi Tom! There is a
 by Vadim Kosoy on Delegative Inverse Reinforcement Learning | 0 likes

Hi Alex! I agree that the
 by Vadim Kosoy on Cooperative Oracles: Stratified Pareto Optima and ... | 0 likes

That is a good question. I
 by Tom Everitt on CIRL Wireheading | 0 likes

 by Tom Everitt on CIRL Wireheading | 0 likes

"The use of an advisor allows
 by Tom Everitt on Delegative Inverse Reinforcement Learning | 0 likes

 by Wei Dai on Current thoughts on Paul Christano's research agen... | 0 likes

Suppose that I, Paul, use a
 by Paul Christiano on Current thoughts on Paul Christano's research agen... | 0 likes

When you wrote "suppose I use
 by Wei Dai on Current thoughts on Paul Christano's research agen... | 0 likes

> but that kind of white-box
 by Paul Christiano on Current thoughts on Paul Christano's research agen... | 0 likes

>Competence can be an
 by Wei Dai on Current thoughts on Paul Christano's research agen... | 0 likes