paper-with-me

홈 › Papers

Context-Scaling versus Task-Scaling in In-Context Learning

2024-10-16 · Amirhesam Abedsoltan, Adityanarayanan Radhakrishnan, Jingfeng Wu, Mikhail Belkin

Transformers exhibit In-Context Learning (ICL), where these models solve new tasks by using examples in the prompt without additional training. In our work, we identify and analyze two key components of ICL: (1) context-scaling, where model performance improves as the number of in-context examples increases and (2) task-scaling, where model performance improves as the number of pre-training tasks increases. While transformers are capable of both context-scaling and task-scaling, we empirically show that standard Multi-Layer Perceptrons (MLPs) with vectorized input are only capable of task-scaling. To understand how transformers are capable of context-scaling, we first propose a significantly simplified transformer architecture without key, query, value weights. We show that it performs ICL comparably to the original GPT-2 model in various statistical learning tasks including linear regression, teacher-student settings. Furthermore, a single block of our simplified transformer can be viewed as data dependent feature map followed by an MLP. This feature map on its own is a powerful predictor that is capable of context-scaling but is not capable of task-scaling. We show empirically that concatenating the output of this feature map with vectorized data as an input to MLPs enables both context-scaling and task-scaling. This finding provides a simple setting to study context and task-scaling for ICL.

📄 PDF Abstract BibTeX arXiv:2410.12783

Code (0)

등록된 구현이 없습니다.

Tasks

In-Context Learning

Methods 이 논문이 사용한 방법론

Refunds@Expedia|||How do I get a full refund from Expedia? “How do I get a full refund from Expedia? How do I get a full refund from Expedia? – Call ☎️ +1-(888) 829 (0881) or +1-805-330-4056 or +1-805-330-4056 for Quick Help &…
Attention 설명 없음
Cosine Annealing Cosine Annealing is a type of learning rate schedule that has the effect of starting with a large learning rate that is relatively rapidly decreased to a minimum value before…
Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Multi-Head Attention 설명 없음
Dropout Dropout is a regularization technique for neural networks that drops a unit (along with connections) at training time with a specified probability $p$ (a common value is…
Layer Normalization Unlike batch normalization, Layer Normalization directly estimates the normalization statistics from the summed inputs…
Linear Warmup With Cosine Annealing Linear Warmup With Cosine Annealing is a learning rate schedule where we increase the learning rate linearly for $n$ updates and then anneal according to a cosine schedule…

Similar Papers 제목 키워드 기반

The Cost of Down-Scaling Language Models: Fact Recall Deteriorates before In-Context Learning

2023-10-07 · Tian Jin, Nolan Clement, Xin Dong, Vaishnavh Nagarajan 외

How does scaling the number of parameters in large language models (LLMs) affect their core capabilities? We study two natural scaling techniques -- weight pruning and simply training a smaller or larger model, which we …

In-Context Learning

Extending Test-Time Scaling: A 3D Perspective with Context, Batch, and Turn

2025-11-18 · Chao Yu, Qixin Tan, Jiaxuan Gao, Shi Yu 외 arxiv

Reasoning reinforcement learning (RL) has recently revealed a new scaling effect: test-time scaling. Thinking models such as R1 and o1 improve their reasoning accuracy at test time as the length of the reasoning context …

Reinforcement Learning

Critical attention scaling in long-context transformers

2025-10-07 · Shi Chen, Zhengjiang Lin, Yury Polyanskiy, Philippe Rigollet arxiv

As large language models scale to longer contexts, attention layers suffer from a fundamental pathology: attention scores collapse toward uniformity as context length $n$ increases, causing tokens to cluster excessively,…

AbbIE: Autoregressive Block-Based Iterative Encoder for Efficient Sequence Modeling

2025-07-11 · Preslav Aleksandrov, Meghdad Kurmanji, Fernando Garcia Redondo, David O'Shea 외 arxiv

We introduce the Autoregressive Block-Based Iterative Encoder (AbbIE), a novel recursive generalization of the encoder-only Transformer architecture, which achieves better perplexity than a standard Transformer and allow…

Two to Five Truths in Non-Negative Matrix Factorization

2023-05-06 · John M. Conroy, Neil P Molino, Brian Baughman, Rod Gomez 외

In this paper, we explore the role of matrix scaling on a matrix of counts when building a topic model using non-negative matrix factorization. We present a scaling inspired by the normalized Laplacian (NL) for graphs th…

Clusteringdocument understandingGraph ClusteringSpectral Graph Clustering+2