Inverse Scaling: When Bigger Isn't Better

By I. R. McKenzie, Alexander Lyzhov, M. Pieler, Alicia Parrish, Aaron Mueller, Ameya Prabhu, Euan McLean, Aaron Kirtland, Alexis Ross, Alisa Liu and 19 more. 2023

235 citations recorded by the source corpus.

Work on scaling laws has found that large language models (LMs) show predictable improvements to overall loss with increased scale (model size, training data, and compute). Here, we present evidence for the claim that LMs may show inverse scaling, or worse task performance with increased scale, e.g., due to flaws in the training objective and data. We present empirical evidence of inverse scaling on 11 datasets collected by running a public contest, the Inverse Scaling Prize, with a substantial prize pool. Through analysis of the datasets, along with other examples found in the literature, we identify four potential causes of inverse scaling: (i) preference to repeat memorized sequences over following in-context instructions, (ii) imitation of undesirable patterns in the training data, (iii) tasks containing an easy distractor task which LMs could focus on, rather than the harder real task, and (iv) correct but misleading few-shot demonstrations of the task. We release the winning datasets at https://inversescaling.com/data to allow for further investigation of inverse scaling. Our tasks have helped drive the discovery of U-shaped and inverted-U scaling trends, where an initial trend reverses, suggesting that scaling trends are less reliable at predicting the behavior of larger-scale models than previously understood. Overall, our results suggest that there are tasks for which increased model scale alone may not lead to progress, and that more careful thought needs to go into the data and objectives for training language models.

Subfields: Artificial Intelligence, Scaling & Learning Theory

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Scaling & Learning Theory

Citations in the corpus

Inverse scaling can become U-shaped (2022, methodology): “The Inverse Scaling Prize was created to identify such tasks for which larger LMs show increasingly undesirable behavior (McKenzie et al., 2023).”
Weak-to-Strong Generalization: Eliciting Strong Capabilities With Weak Supervision (2023, background): “Surprisingly, we find inverse scaling (McKenzie et al., 2023): larger student models consistently agree less with the errors of the supervisor than smaller student models, despite being trained to imitate the supervisor, not using early stopping, and having larger capacity than smaller student…”
Reasoning or Reciting? Exploring the Capabilities and Limitations of Language Models Through Counterfactual Tasks (2023, result): “Nevertheless, McKenzie et al. (2023), who found a similar trend but with respect to pretraining FLOPs and termed it “inverse scaling,” also provided a memorization-based explanation: they observed that when a task contradicts with pretraining texts, similar to how our counterfactual conditions…”
Jailbreak in pieces: Compositional Adversarial Attacks on Multi-Modal Language Models (2023, unknown): “This could serve as an example of the “Inverse-Scaling” phenomenon introduced by McKenzie et al. (2023), wherein larger models tend to excel in following instructions and managing more extensive contexts leading to expanded attack surfaces as also studied deeply by Wei et al. (2023).”
Measuring Faithfulness in Chain-of-Thought Reasoning (2023, background): “As a result, we hypothesize that, for reasoning faithfulness, larger models may behave worse than smaller ones (“inverse scaling”; McKenzie et al., 2023); smaller models may, for some tasks, beneﬁt more from CoT, potentially leading them to rely more on CoT.”