Like technically yes, it shows that there is an internal optimization process that is running in the networks, but much of the meat of optimization such as instrumental convergence/power-seeking depends the structure of the function one is optimizing over.
If the function is not consequentialist—if it doesn’t attempt to compute the real-world consequences of different outputs and grade things based on those consequences—then much of the discussion about optimizers does not apply.
Well, no, that’s not the definition of optimizer in the mesa-optimization post! Evan gives the following definition of an optimizer:
A system is an optimizer if it is internally searching through a search space (consisting of possible outputs, policies, plans, strategies, or similar) looking for those elements that score high according to some objective function that is explicitly represented within the system
And the following definition of a mesa-optimizer:
Mesa-optimization occurs when a base optimizer (in searching for algorithms to solve some problem) finds a model that is itself an optimizer, which we will call a mesa-optimizer.
In this paper, the authors show that transformers gradient descent (an optimization algorithm) to optimize a particular objective (ℓ2 loss). (This is very similar to the outer optimization loop that’s altering the network parameters.) So the way davidad uses the word “mesa-optimization” is consistent with prior work.
I also think that it’s pretty bad to claim that something is only an optimizer if it’s a power-seeking consequentialist agent. For example, this would imply that the outer loop that produces neural network policies (by gradient descent on network parameters) is not an optimizer!
Of course, I agree that it’s not the case that these transformers are power-seeking consequentialist agents. And so this paper doesn’t provide direct evidence that transformers contain power-seeking consequentialist agents (except for people who disbelieved in power-seeking consequentialist agents because they thought NNs are basically incapable of implementing any optimizer whatsoever).
some objective function that is explicitly represented within the system
but that is not the case here.
There is a fundamental difference between
Programs that implement the computation of taking the derivative. (f→f′, or perhaps f,x→f′(x).)
Programs that implement some particular function g, which happens to be the derivative of some other function. (x→g(x), where it so happens that g=F′ for some F.)
The transformers in this paper are programs of the 2nd type. They don’t contain any logic about taking the gradient of an arbitrary function, and one couldn’t “retarget” them toward L1 loss or some other function.
(One could probably construct similar layers that implement the gradient step for L1, but they’d again be programs of the 2nd type, just with a different hardcoded g.)
Calling something like this an optimizer strikes me as vacuous: if you don’t require the ability to adapt to a change of objective function, you can always take any program and say it’s “optimizing” some function. Just pick a function that’s maximal when you do whatever it is that the program does.
It’s not vacuous to say that the transformers in the paper “implement gradient descent,” as long as one means they “implement [gradient descent on L2 loss]” rather than “implement [gradient descent] on [L2 loss].” They don’t implement general gradient descent, but happen to coincide with the gradient step for L2 loss.
If in-content learning in real transformers involves figuring out the objective function from the context, then this result cannot explain it. If we assume some fixed objective function (perhaps LM loss itself?) and ask whether the model might be doing gradient steps on this function internally, then these results are relevant.
I think the claim that an optimizer is a retargetable search process makes a lot of sense* and I’ve edited the post to link to this clarification.
That being said, I’m still confused about the details.
Suppose that I do a goal-conditioned version of the paper, where (hypothetically) I exhibit a transformer circuit that, conditioned on some prompt or the other, was able to alternate between performing gradient descent on three types of objectives (say, L1, L2, L\infty) -- would this suffice? How about if, instead, there wasn’t any prompt that let me switch between three types of objectives, but there was a parameter inside of the neural network that I could change that causes the circuit to optimize different objectives? How much of the circuit do I have to change before it becomes a new circuit instead of retargeting the optimizer?
I guess part of answer to these questions might look like, “there might not be a clear cutoff, in the same sense that there’s not a clear cutoff for most other definitions that we use in AI alignment (‘agent’ or ‘deceptive alignment’ for example)”, while another part might be “this is left for future work”.
We now characterize goal misgeneralization. Intuitively, goal misgeneralization occurs when we learn a function fθbad that has robust capabilities but pursues an undesired goal.
It is quite challenging to define what a “capability” is in the context of neural networks. We provide a provisional definition following Chen et al. [11]. We say that the model is capable of some task X in setting Y if it can be quickly tuned to perform task X well in setting Y (relative to learning X from scratch). For example, tuning could be done by prompt engineering or by fine-tuning on a small quantity of data [52].
I am aware of the definition of a mesa-optimizer. I’m just saying that this definition doesn’t catch much of the danger or exciting stuff, since it’s e.g. not so important if the model contains something that optimizes over e.g. L(x)=(x2−2)2. The danger comes when the inner objective is more consequentialist than that.
I also think that it’s pretty bad to claim that something is only an optimizer if it’s a power-seeking consequentialist agent. For example, this would imply that the outer loop that produces neural network policies (by gradient descent on network parameters) is not an optimizer!
If the outer loop isn’t connected to some mechanism that evaluates the consequences of different policies in the real world, then you are probably training something that mimics the prespecified training data rather than searching for novel powerful policies. This isn’t useless—the explosive growth of statistical models for various purposes proves so much—but it is unlikely to be dangerous unless coupled with some other process that handles the optimization.
So I’m not saying these things aren’t optimizers, but much of the worry about mesa-optimizers is about consequentialist optimizers, not generic optimizers. Most functions are perfectly safe to optimize.
Yes, it’s worth pulling out that the mesa-optimizers demonstrated here are not consequentialists, they are optimizing the goodness of fit of an internal representation to in-context data.
The role this plays in arguments about deceptive alignment is that it neutralizes the claim that “it’s probably not a realistically efficient or effective or inductive-bias-favoured structure to actually learn an internal optimization algorithm”. Arguments like “it’s not inductive-bias-favoured for mesa-optimizers to be consequentialists instead of maximizing purely epistemic utility” remain.
Although I predict someone will find consequentialist mesa-optimizers in Decision Transformers, that has not (to my knowledge) actually been seen yet.
I don’t think so.
Like technically yes, it shows that there is an internal optimization process that is running in the networks, but much of the meat of optimization such as instrumental convergence/power-seeking depends the structure of the function one is optimizing over.
If the function is not consequentialist—if it doesn’t attempt to compute the real-world consequences of different outputs and grade things based on those consequences—then much of the discussion about optimizers does not apply.
Well, no, that’s not the definition of optimizer in the mesa-optimization post! Evan gives the following definition of an optimizer:
And the following definition of a mesa-optimizer:
In this paper, the authors show that transformers gradient descent (an optimization algorithm) to optimize a particular objective (ℓ2 loss). (This is very similar to the outer optimization loop that’s altering the network parameters.) So the way davidad uses the word “mesa-optimization” is consistent with prior work.
I also think that it’s pretty bad to claim that something is only an optimizer if it’s a power-seeking consequentialist agent. For example, this would imply that the outer loop that produces neural network policies (by gradient descent on network parameters) is not an optimizer!
Of course, I agree that it’s not the case that these transformers are power-seeking consequentialist agents. And so this paper doesn’t provide direct evidence that transformers contain power-seeking consequentialist agents (except for people who disbelieved in power-seeking consequentialist agents because they thought NNs are basically incapable of implementing any optimizer whatsoever).
That definition of “optimizer” requires
but that is not the case here.
There is a fundamental difference between
Programs that implement the computation of taking the derivative. (f→f′, or perhaps f,x→f′(x).)
Programs that implement some particular function g, which happens to be the derivative of some other function. (x→g(x), where it so happens that g=F′ for some F.)
The transformers in this paper are programs of the 2nd type. They don’t contain any logic about taking the gradient of an arbitrary function, and one couldn’t “retarget” them toward L1 loss or some other function.
(One could probably construct similar layers that implement the gradient step for L1, but they’d again be programs of the 2nd type, just with a different hardcoded g.)
Calling something like this an optimizer strikes me as vacuous: if you don’t require the ability to adapt to a change of objective function, you can always take any program and say it’s “optimizing” some function. Just pick a function that’s maximal when you do whatever it is that the program does.
It’s not vacuous to say that the transformers in the paper “implement gradient descent,” as long as one means they “implement [gradient descent on L2 loss]” rather than “implement [gradient descent] on [L2 loss].” They don’t implement general gradient descent, but happen to coincide with the gradient step for L2 loss.
If in-content learning in real transformers involves figuring out the objective function from the context, then this result cannot explain it. If we assume some fixed objective function (perhaps LM loss itself?) and ask whether the model might be doing gradient steps on this function internally, then these results are relevant.
I think the claim that an optimizer is a retargetable search process makes a lot of sense* and I’ve edited the post to link to this clarification.
That being said, I’m still confused about the details.
Suppose that I do a goal-conditioned version of the paper, where (hypothetically) I exhibit a transformer circuit that, conditioned on some prompt or the other, was able to alternate between performing gradient descent on three types of objectives (say, L1, L2, L\infty) -- would this suffice? How about if, instead, there wasn’t any prompt that let me switch between three types of objectives, but there was a parameter inside of the neural network that I could change that causes the circuit to optimize different objectives? How much of the circuit do I have to change before it becomes a new circuit instead of retargeting the optimizer?
I guess part of answer to these questions might look like, “there might not be a clear cutoff, in the same sense that there’s not a clear cutoff for most other definitions that we use in AI alignment (‘agent’ or ‘deceptive alignment’ for example)”, while another part might be “this is left for future work”.
*This is also similar to the definition used for inner misalignment in Shah et al’s Goal Misgeneralization paper:
I am aware of the definition of a mesa-optimizer. I’m just saying that this definition doesn’t catch much of the danger or exciting stuff, since it’s e.g. not so important if the model contains something that optimizes over e.g. L(x)=(x2−2)2. The danger comes when the inner objective is more consequentialist than that.
If the outer loop isn’t connected to some mechanism that evaluates the consequences of different policies in the real world, then you are probably training something that mimics the prespecified training data rather than searching for novel powerful policies. This isn’t useless—the explosive growth of statistical models for various purposes proves so much—but it is unlikely to be dangerous unless coupled with some other process that handles the optimization.
So I’m not saying these things aren’t optimizers, but much of the worry about mesa-optimizers is about consequentialist optimizers, not generic optimizers. Most functions are perfectly safe to optimize.
Yes, it’s worth pulling out that the mesa-optimizers demonstrated here are not consequentialists, they are optimizing the goodness of fit of an internal representation to in-context data.
The role this plays in arguments about deceptive alignment is that it neutralizes the claim that “it’s probably not a realistically efficient or effective or inductive-bias-favoured structure to actually learn an internal optimization algorithm”. Arguments like “it’s not inductive-bias-favoured for mesa-optimizers to be consequentialists instead of maximizing purely epistemic utility” remain.
Although I predict someone will find consequentialist mesa-optimizers in Decision Transformers, that has not (to my knowledge) actually been seen yet.