U-shaped and Inverted-U Scaling behind Emergent Abilities of Large Language Models

• URL: http://arxiv.org/abs/2410.01692v1
• Date: Wed, 2 Oct 2024 16:03:49 GMT
• Title: U-shaped and Inverted-U Scaling behind Emergent Abilities of Large Language Models
• Authors: Tung-Yu Wu, Pei-Yu Lo,
• Abstract summary: Large language models (LLMs) have been shown to exhibit emergent abilities in some downstream tasks. We observe U-shaped scaling for hard questions, and inverted-U scaling followed by steady improvement for easy questions. We propose a simple yet effective pipeline, called Slice-and-Sandwich, to predict both the emergence threshold and model performance beyond the threshold.
• Score: 1.14179290793997
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large language models (LLMs) have been shown to exhibit emergent abilities in some downstream tasks, where performance seems to stagnate at first and then improve sharply and unpredictably with scale beyond a threshold. By dividing questions in the datasets according to difficulty level by average performance, we observe U-shaped scaling for hard questions, and inverted-U scaling followed by steady improvement for easy questions. Moreover, the emergence threshold roughly coincides with the point at which performance on easy questions reverts from inverse scaling to standard scaling. Capitalizing on the observable though opposing scaling trend on easy and hard questions, we propose a simple yet effective pipeline, called Slice-and-Sandwich, to predict both the emergence threshold and model performance beyond the threshold.

Related papers

• Scale-Invariant Learning-to-Rank [0.0]
  At Expedia, learning-to-rank models play a key role in sorting and presenting information more relevant to users. A major challenge in deploying these models is ensuring consistent feature scaling between training and production data. We introduce a scale-invariant LTR framework which combines a deep and a wide neural network to mathematically guarantee scale-invariance in the model at both training and prediction time. We evaluate our framework in simulated real-world scenarios with injected feature scale issues by perturbing the test set at prediction time, and show that even with inconsistent train-test scaling, using framework achieves better performance than
  arXiv  Detail & Related papers  (2024-10-02T19:05:12Z)
• Observational Scaling Laws and the Predictability of Language Model Performance [51.2336010244645]
  We propose an observational approach that bypasses model training and instead builds scaling laws from 100 publically available models. We show that several emergent phenomena follow a smooth, sigmoidal behavior and are predictable from small models. We show how to predict the impact of post-training interventions like Chain-of-Thought and Self-Consistency as language model capabilities continue to improve.
  arXiv  Detail & Related papers  (2024-05-17T17:49:44Z)
• Predicting Emergent Abilities with Infinite Resolution Evaluation [85.89911520190711]
  We introduce PassUntil, an evaluation strategy with theoretically infinite resolution, through massive sampling in the decoding phase. We predict the performance of the 2.4B model on code generation with merely 0.05% deviation before training starts. We identify a kind of accelerated emergence whose scaling curve cannot be fitted by standard scaling law function.
  arXiv  Detail & Related papers  (2023-10-05T02:35:00Z)
• Inverse Scaling: When Bigger Isn't Better [80.42834197416444]
  Large language models (LMs) show predictable improvements to overall loss with increased scale. We present evidence for the claim that LMs may show inverse scaling, or worse task performance with increased scale.
  arXiv  Detail & Related papers  (2023-06-15T20:11:23Z)
• Beyond Positive Scaling: How Negation Impacts Scaling Trends of Language Models [92.11542797811461]
  We introduce NeQA, a dataset consisting of questions with negation. We show that this task can exhibit inverse scaling, U-shaped scaling, or positive scaling. We find that task 1 has linear scaling, while task 2 has sigmoid-shaped scaling with an emergent transition point.
  arXiv  Detail & Related papers  (2023-05-27T00:07:17Z)
• Emergent inabilities? Inverse scaling over the course of pretraining [0.6091702876917281]
  We investigate whether, over the course of training, the performance of language models at specific tasks can decrease while general performance remains high. We find that for two tasks from the Inverse Scaling Challenge - quote-repetition and redefine-math - this is indeed the case. This highlights the importance of testing model performance at all relevant benchmarks any time they are trained on additional data, even if their overall performance improves.
  arXiv  Detail & Related papers  (2023-05-24T03:42:43Z)
• Inverse scaling can become U-shaped [126.64521446943155]
  Scaling up language models has been empirically shown to improve performance on a wide range of downstream tasks. This paper takes a closer look at these inverse scaling tasks. We evaluate models of up to 540B parameters, trained on five times more compute than those evaluated in the Inverse Scaling Prize.
  arXiv  Detail & Related papers  (2022-11-03T17:26:44Z)
• Dynamic Scale Training for Object Detection [111.33112051962514]
  We propose a Dynamic Scale Training paradigm (abbreviated as DST) to mitigate scale variation challenge in object detection. Experimental results demonstrate the efficacy of our proposed DST towards scale variation handling. It does not introduce inference overhead and could serve as a free lunch for general detection configurations.
  arXiv  Detail & Related papers  (2020-04-26T16:48:17Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.