原始对话
Dwarkesh Podcast · · 1:37:01
John Schulman discusses uncertainty about generalization, improving experimental methods, and the role of realistic prompt distributions in model distillation.
简短、标注来源的段落，并附有可验证上下文。完整对话保留在其发布者处。
原始摘录
The whole field relies a lot on generalization and it’s very hard to predict when you’re going to get generalization, or when you’re going to get some kind of out-of-distribution generalization
此片段未附更多上下文，请阅读原始访谈。
原始摘录
If you think hard enough, you probably could have expected some of these things beforehand. There is probably some very clever way to do a small-scale experiment that’ll let you build the theory that then will generalize to the large-scale experiment. So I would expect that we’re nowhere near the ceiling of how well you can do research
此片段未附更多上下文，请阅读原始访谈。
原始摘录
If you have a really good realistic prompt distribution for distillation, you can match the big model really well. But if you only have this distribution of easily verifiable tasks, then you can match the big model on all the benchmarks, but you do worse on this broader distribution
此片段未附更多上下文，请阅读原始访谈。
这些观点均关联原始来源。转述已明确标注，不作为逐字原话展示。
打开转录或来源材料 （在新标签页中打开）报告问题
nafyi.