%0 Conference Paper %1 schaeffer2023emergent %A Schaeffer, Rylan %A Miranda, Brando %A Koyejo, Sanmi %B Proceedings of the 37th International Conference on Neural Information Processing Systems %C Red Hook, NY, USA %D 2023 %I Curran Associates Inc. %K ai emergence evaluation learning llm machine math metric ml %P 55565--55581 %T Are emergent abilities of large language models a mirage? %U https://proceedings.neurips.cc/paper_files/paper/2023/hash/adc98a266f45005c403b8311ca7e8bd7-Abstract-Conference.html %X Recent work claims that large language models display emergent abilities: abilities not present in smaller-scale models that are present in larger-scale models. What makes emergent abilities intriguing is two-fold: their sharpness, transitioning seemingly instantaneously from not present to present, and their unpredictability, appearing at seemingly unforeseeable model scales. Here, we present an alternative explanation for emergent abilities: for a particular task and model family, when analyzing fixed model outputs, emergent abilities appear due to the researcher's choice of metric rather than due to fundamental changes in models with scale. Specifically, nonlinear or discontinuous metrics produce seemingly emergent abilities, whereas linear or continuous metrics produce smooth, continuous, predictable changes in model performance. We present our alternative explanation in a simple mathematical model, then test it in three complementary ways: we (1) make, test and confirm three predictions on the effect of metric choice using the InstructGPT/GPT-3 family on tasks with claimed emergent abilities; (2) make, test and confirm two predictions about metric choices in a meta-analysis of emergent abilities on the Beyond the Imitation Game Benchmark (BIG-Bench); and (3) show how to choose metrics to produce never-before-seen seemingly emergent abilities in multiple vision tasks across diverse deep network architectures. Via all three analyses, we provide evidence that emergent abilities disappear with different metrics or with better statistics, and may not be a fundamental property of scaling AI models.

@inproceedings{schaeffer2023emergent, abstract = {Recent work claims that large language models display emergent abilities: abilities not present in smaller-scale models that are present in larger-scale models. What makes emergent abilities intriguing is two-fold: their sharpness, transitioning seemingly instantaneously from not present to present, and their unpredictability, appearing at seemingly unforeseeable model scales. Here, we present an alternative explanation for emergent abilities: for a particular task and model family, when analyzing fixed model outputs, emergent abilities appear due to the researcher's choice of metric rather than due to fundamental changes in models with scale. Specifically, nonlinear or discontinuous metrics produce seemingly emergent abilities, whereas linear or continuous metrics produce smooth, continuous, predictable changes in model performance. We present our alternative explanation in a simple mathematical model, then test it in three complementary ways: we (1) make, test and confirm three predictions on the effect of metric choice using the InstructGPT/GPT-3 family on tasks with claimed emergent abilities; (2) make, test and confirm two predictions about metric choices in a meta-analysis of emergent abilities on the Beyond the Imitation Game Benchmark (BIG-Bench); and (3) show how to choose metrics to produce never-before-seen seemingly emergent abilities in multiple vision tasks across diverse deep network architectures. Via all three analyses, we provide evidence that emergent abilities disappear with different metrics or with better statistics, and may not be a fundamental property of scaling AI models.}, added-at = {2025-02-26T15:00:45.000+0100}, address = {Red Hook, NY, USA}, author = {Schaeffer, Rylan and Miranda, Brando and Koyejo, Sanmi}, biburl = {https://www.bibsonomy.org/bibtex/22a292361c3414cc8b4d4c487602f6ecf/jaeschke}, booktitle = {Proceedings of the 37th International Conference on Neural Information Processing Systems}, day = 10, description = {Are emergent abilities of large language models a mirage? | Proceedings of the 37th International Conference on Neural Information Processing Systems}, interhash = {33ade2acc01952ff8fc9745fdd93ba04}, intrahash = {2a292361c3414cc8b4d4c487602f6ecf}, keywords = {ai emergence evaluation learning llm machine math metric ml}, location = {New Orleans, LA, USA}, month = {12}, pages = {55565--55581}, publisher = {Curran Associates Inc.}, series = {NIPS '23}, timestamp = {2025-02-26T15:02:45.000+0100}, title = {Are emergent abilities of large language models a mirage?}, url = {https://proceedings.neurips.cc/paper_files/paper/2023/hash/adc98a266f45005c403b8311ca7e8bd7-Abstract-Conference.html}, year = 2023 }