Mark Xu

I do alignment research at the Alignment Research Center. Learn more about me at markxu.com/about

Sequences

Intermittent Distllations

Wiki Contributions

Comments

Prizes for ELK proposals

We generally assume that we can construct questions sufficiently well that there's only one unambiguous interpretation. We also generally assume that the predictor "knows" which world it's in because it can predict how humans would respond to hypothetical questions about various situations involving diamonds and sensors and that humans would say in theory Q1 and Q2 could be different.

More concretely, our standard for judging proposals is exhibiting an unambiguous failure. If it was plausible you asked the wrong question, or the AI didn't know what you meant by the question, then the failure exhibited would be ambiguous. If humans are unable to clarify between two possible interpretations of their question, then the failure would be ambiguous.

Alex Ray's Shortform

I think we would be trying to elicit obfuscated knowledge in ELK. In our examples, you can imagine that the predictor's Bayes net works "just because", so an argument that is convincing to a human for why the diamond in the room has to be arguing that the Bayes net is a good explanation of reality + arguing that it implies the diamond is in the room, which is the sort of "obfuscated" knowledge that debate can't really handle.

Prizes for ELK proposals

The dataset is generated with the human bayes net, so it's sufficient to map to the human bayes net. There is, of course, an infinite set of "human" simulators that use slightly different bayes nets that give the same answers on the training set.

Prizes for ELK proposals

Does this mean that the method needs to work for ~arbitrary architectures, and that the solution must use substantially the same architecture as the original?

Yes, approximately. If you can do it for only e.g. transformers, but not other things, that would be interesting.

Does this mean that it must be able to deal with a broad variety of questions, so that we cannot simply sit down and think about how to optimize the model for getting a single question (e.g. "Where is the diamond?") right?

Yes, approximately. Thinking about how to get one question right might be a productive way to do research. However, if you have a strategy for answering 1 question right, it should also work for other questions.

Prizes for ELK proposals

We generally imagine that it’s impossible to map the predictors net directly to an answer because the predictor is thinking in terms of different concepts, so it has to map to the humans nodes first in order to answer human questions about diamonds and such.

Prizes for ELK proposals

The SmartFabricator seems basically the same. In the robber example, you might imagine the SmartVault is the one that puts up the screen to conceal the fact that it let the diamond get stolen.

Prizes for ELK proposals

A different way of phrasing Ajeya's response, which I think is roughly accurate, is that if you have a reporter that gives consistent answers to questions, you've learned a fact about the predictor, namely "the predictor was such that when it was paired with this reporter it gave consistent answers to questions." if there were 8 predictor for which this fact was true then "it's the [7th] predictor such that when it was paired with this reporter it gave consistent answers to questions" is enough information to uniquely determine the reporter, e.g. the previous fact + 3 additional bits was enough. if the predictor was 1000 bits, the fact that it was consistent with a reporter "saved" you 997 bits, compressing the predictor into 3 bits.

The hope is that maybe the honest reporter "depends" on larger parts of the predictor's reasoning, so less predictors are consistent with it, so the fact that a predictor is consistent with the honest reporter allows you to compress the predictor more. As such, searching for reporters that most compressed the predictor would prefer the honest reporter. However, the best way for a reporter to compress a predictor is to simply memorize the entire thing, so if the predictor is simple enough and the gap between the complexity of the human-imitator and the direct translator is large enough, then the human-imitator+memorized predictor is the simplest thing that maximally compresses the predictor.

Prizes for ELK proposals

There is a distinction between the way that the predictor is reasoning and the way that the reporter works. Generally, we imagine that that the predictor is trained the same way the "unaligned benchmark" we're trying to compare to is trained, and the reporter is the thing that we add onto that to "align" it (perhaps by only training another head on the model, perhaps by finetuning). Hopefully, the cost of training the reporter is small compared to the cost of the predictor (maybe like 10% or something)

In this frame, doing anything to train the way the predictor is trained results in a big competitiveness hit, e.g. forcing the predictor to use the same ontology as a human is potentially going to prevent it from using concepts that make reasoning much more efficient. However, training the reporter in a different way, e.g. doubling the cost of training the reporter, only takes you from 10% of the predictor to 20%, which not that bad of a competitiveness hit (assuming that the human imitator takes 10% of the cost of the original predictor to train).

In summary, competitiveness for ELK proposals primarily means that you can't change the way the predictor was trained. We are already assuming/hoping the reporter is much cheaper to train than the predictor, so making the reporter harder to train results in a much smaller competitiveness hit.

Load More