Do pretrained Transformers Learn In-Context by Gradient Descent?

• URL: http://arxiv.org/abs/2310.08540v5
• Date: Mon, 3 Jun 2024 04:18:11 GMT
• Title: Do pretrained Transformers Learn In-Context by Gradient Descent?
• Authors: Lingfeng Shen, Aayush Mishra, Daniel Khashabi,
• Abstract summary: In this paper, we investigate the emergence of In-Context Learning (ICL) in language models pre-trained on natural data (LLaMa-7B) We find that ICL and Gradient Descent (GD) modify the output distribution of language models differently. These results indicate that emphthe equivalence between ICL and GD remains an open hypothesis and calls for further studies.
• Score: 21.23795112800977
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The emergence of In-Context Learning (ICL) in LLMs remains a remarkable phenomenon that is partially understood. To explain ICL, recent studies have created theoretical connections to Gradient Descent (GD). We ask, do such connections hold up in actual pre-trained language models? We highlight the limiting assumptions in prior works that make their setup considerably different from the practical setup in which language models are trained. For example, their experimental verification uses \emph{ICL objective} (training models explicitly for ICL), which differs from the emergent ICL in the wild. Furthermore, the theoretical hand-constructed weights used in these studies have properties that don't match those of real LLMs. We also look for evidence in real models. We observe that ICL and GD have different sensitivity to the order in which they observe demonstrations. Finally, we probe and compare the ICL vs. GD hypothesis in a natural setting. We conduct comprehensive empirical analyses on language models pre-trained on natural data (LLaMa-7B). Our comparisons of three performance metrics highlight the inconsistent behavior of ICL and GD as a function of various factors such as datasets, models, and the number of demonstrations. We observe that ICL and GD modify the output distribution of language models differently. These results indicate that \emph{the equivalence between ICL and GD remains an open hypothesis} and calls for further studies.

Related papers

• Bayesian scaling laws for in-context learning [72.17734205418502]
  In-context learning (ICL) is a powerful technique for getting language models to perform complex tasks with no training updates. We show that ICL approximates a Bayesian learner and develop a family of novel Bayesian scaling laws for ICL.
  arXiv  Detail & Related papers  (2024-10-21T21:45:22Z)
• Probing Causality Manipulation of Large Language Models [12.46951388060595]
  Large language models (LLMs) have shown various ability on natural language processing, including problems about causality. This paper proposes a novel approach to probe causality manipulation hierarchically, by providing different shortcuts to models and observe behaviors.
  arXiv  Detail & Related papers  (2024-08-26T16:00:41Z)
• Is In-Context Learning a Type of Gradient-Based Learning? Evidence from the Inverse Frequency Effect in Structural Priming [6.408190458163885]
  Large language models (LLMs) have shown the emergent capability of in-context learning (ICL) We introduce a new way of diagnosing whether ICL is functionally equivalent to gradient-based learning.
  arXiv  Detail & Related papers  (2024-06-26T17:06:41Z)
• What Do Language Models Learn in Context? The Structured Task Hypothesis [89.65045443150889]
  Large language models (LLMs) learn a novel task from in-context examples presented in a demonstration, termed in-context learning (ICL) One popular hypothesis explains ICL by task selection. Another popular hypothesis is that ICL is a form of meta-learning, i.e., the models learn a learning algorithm at pre-training time and apply it to the demonstration.
  arXiv  Detail & Related papers  (2024-06-06T16:15:34Z)
• The Strong Pull of Prior Knowledge in Large Language Models and Its Impact on Emotion Recognition [74.04775677110179]
  In-context Learning (ICL) has emerged as a powerful paradigm for performing natural language tasks with Large Language Models (LLM) We show that LLMs have strong yet inconsistent priors in emotion recognition that ossify their predictions. Our results suggest that caution is needed when using ICL with larger LLMs for affect-centered tasks outside their pre-training domain.
  arXiv  Detail & Related papers  (2024-03-25T19:07:32Z)
• Exploring the Relationship between In-Context Learning and Instruction Tuning [18.186126518966017]
  In-Context Learning (ICL) and Instruction Tuning (IT) are two primary paradigms of adopting Large Language Models to downstream applications. In ICL, a set of demonstrations are provided at inference time but the LLM's parameters are not updated. In IT, a set of demonstrations are used to tune LLM's parameters in training time but no demonstrations are used at inference time.
  arXiv  Detail & Related papers  (2023-11-17T07:40:46Z)
• Dynamic Demonstrations Controller for In-Context Learning [51.3439660534631]
  In-Context Learning (ICL) is a new paradigm for natural language processing (NLP), where a large language model observes a small number of demonstrations and a test instance as its input. Previous studies have revealed that ICL is sensitive to the selection and the ordering of demonstrations. We propose a Dynamic Demonstrations Controller (D$2$Controller), which can improve the ICL performance by adjusting the number of demonstrations.
  arXiv  Detail & Related papers  (2023-09-30T14:04:22Z)
• Investigating the Learning Behaviour of In-context Learning: A Comparison with Supervised Learning [67.25698169440818]
  Large language models (LLMs) have shown remarkable capacity for in-context learning (ICL) We train the same LLMs with the same demonstration examples via ICL and supervised learning (SL), respectively, and investigate their performance under label perturbations. First, we find that gold labels have significant impacts on the downstream in-context performance, especially for large language models. Second, when comparing with SL, we show empirically that ICL is less sensitive to label perturbations than SL, and ICL gradually attains comparable performance to SL as the model size increases.
  arXiv  Detail & Related papers  (2023-07-28T09:03:19Z)
• Explaining Emergent In-Context Learning as Kernel Regression [61.57151500616111]
  Large language models (LLMs) have initiated a paradigm shift in transfer learning. In this paper, we investigate the reason why a transformer-based language model can accomplish in-context learning after pre-training. We find that during ICL, the attention and hidden features in LLMs match the behaviors of a kernel regression.
  arXiv  Detail & Related papers  (2023-05-22T06:45:02Z)

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.