It would also be very useful to build some GPT feature “visualization” tools ASAP.
Do you have anything more specific in mind? I see the Image Feature Visualization tool, but in my mind it’s basically doing exactly what you’re already doing by comparing GPT-2 and GPT-3 snippets.
No, the closest analogue of comparing text snippets is staring at image completions, which is not nearly as informative as being able to go neuron-by-neuron or layer-by-layer and get a sense of the concepts at each level.
Do you have anything more specific in mind? I see the Image Feature Visualization tool, but in my mind it’s basically doing exactly what you’re already doing by comparing GPT-2 and GPT-3 snippets.
No, the closest analogue of comparing text snippets is staring at image completions, which is not nearly as informative as being able to go neuron-by-neuron or layer-by-layer and get a sense of the concepts at each level.