paper-with-me

Papers

Revisiting Dynamic Evaluation: Online Adaptation for Large Language Models

2024-03-03 · Amal Rannen-Triki, Jorg Bornschein, Razvan Pascanu, Marcus Hutter, Andras György, Alexandre Galashov, Yee Whye Teh, Michalis K. Titsias

We consider the problem of online fine tuning the parameters of a language model at test time, also known as dynamic evaluation. While it is generally known that this approach improves the overall predictive performance, especially when considering distributional shift between training and evaluation data, we here emphasize the perspective that online adaptation turns parameters into temporally changing states and provides a form of context-length extension with memory in weights, more in line with the concept of memory in neuroscience. We pay particular attention to the speed of adaptation (in terms of sample efficiency),sensitivity to the overall distributional drift, and the computational overhead for performing gradient computations and parameter updates. Our empirical study provides insights on when online adaptation is particularly interesting. We highlight that with online adaptation the conceptual distinction between in-context learning and fine tuning blurs: both are methods to condition the model on previously observed tokens.

📄 PDF Abstract BibTeX arXiv:2403.01518

Code (0)

등록된 구현이 없습니다.

Tasks

In-Context LearningLanguage ModelingLanguage Modelling

Methods 이 논문이 사용한 방법론

SPEED The monocular depth estimation (MDE) is the task of estimating depth from a single frame. This information is an essential knowledge in many computer vision tasks such as scene…

Similar Papers 제목 키워드 기반

Cumulative Learning Rate Adaptation: Revisiting Path-Based Schedules for SGD and Adam

2025-08-07 · Asma Atamna, Tom Maus, Fabian Kievelitz, Tobias Glasmachers arxiv

The learning rate is a crucial hyperparameter in deep learning, with its ideal value depending on the problem and potentially changing during training. In this paper, we investigate the practical utility of adaptive lear…

UDA-Bench: Revisiting Common Assumptions in Unsupervised Domain Adaptation Using a Standardized Framework

2024-09-23 · Tarun Kalluri, Sreyas Ravichandran, Manmohan Chandraker

In this work, we take a deeper look into the diverse factors that influence the efficacy of modern unsupervised domain adaptation (UDA) methods using a large-scale, controlled empirical study. To facilitate our analysis,…

Domain AdaptationUnsupervised Domain Adaptation

Revisiting the Old Kitchen Sink: Do we Need Sentiment Domain Adaptation?

2013-09-01 · RANLP 2013 9 · Riham Mansour, Nesma Refaei, Michael Gamon, Ahmed Abdul-Hamid 외
Domain Adaptation

Agile Online Model Selection: Resolving Adaptation Lag via Safeguarded Large Learning Rates

2026-05-26 · Kei Takemura, Ryuta Matsuno, Keita Sakuma arxiv

Maintaining predictive accuracy in non-stationary environments requires online model selection to adapt autonomously to unknown distribution shifts. However, existing tuning-free algorithms face a fundamental trade-off b…

DEXTER-LLM: Dynamic and Explainable Coordination of Multi-Robot Systems in Unknown Environments via Large Language Models

2025-08-20 · Yuxiao Zhu, Junfeng Chen, Xintong Zhang, Meng Guo 외 arxiv

Online coordination of multi-robot systems in open and unknown environments faces significant challenges, particularly when semantic features detected during operation dynamically trigger new tasks. Recent large language…