English
Related papers

Related papers: Context Distillation as Latent Memory Management

200 papers

This paper studies the problem of pre-training for small models, which is essential for many mobile devices. Current state-of-the-art methods on this problem transfer the representational knowledge of a large network (as a Teacher) into a…

Computer Vision and Pattern Recognition · Computer Science 2024-06-18 Mingsheng Li , Lin Zhang , Mingzhen Zhu , Zilong Huang , Gang Yu , Jiayuan Fan , Tao Chen

Current long chain-of-thought (long-CoT) models excel at mathematical reasoning but rely on slow and error-prone natural language traces. Tool-augmented agents address arithmetic via code execution, but often falter on complex logical…

Computation and Language · Computer Science 2025-09-03 Weihua Du , Pranjal Aggarwal , Sean Welleck , Yiming Yang

Knowledge distillation between machine learning models has opened many new avenues for parameter count reduction, performance improvements, or amortizing training time when changing architectures between the teacher and student network. In…

Machine Learning · Computer Science 2020-11-24 Jonathan Raiman

Long input sequences are central to in-context learning, document understanding, and multi-step reasoning of Large Language Models (LLMs). However, the quadratic attention cost of Transformers makes inference memory-intensive and slow.…

Computation and Language · Computer Science 2026-02-19 Rujikorn Charakorn , Edoardo Cetin , Shinnosuke Uesaka , Robert Tjarko Lange

Diffusion Language Models (DLMs) have recently achieved strong results in text generation. However, their multi-step sampling leads to slow inference, limiting practical use. To address this, we extend Inverse Distillation, a technique…

Machine Learning · Computer Science 2026-02-24 David Li , Nikita Gushchin , Dmitry Abulkhanov , Eric Moulines , Ivan Oseledets , Maxim Panov , Alexander Korotin

This paper introduces a novel approach for efficiently distilling LLMs into smaller, application-specific models, significantly reducing operational costs and manual labor. Addressing the challenge of deploying computationally intensive…

Computation and Language · Computer Science 2024-03-26 Lukas Vöge , Vincent Gurgul , Stefan Lessmann

Current language models rely on static vocabularies determined at pretraining time, which can lead to decreased performance and increased computational cost for domains underrepresented in the original vocabulary. New tokens can be added to…

Computation and Language · Computer Science 2026-03-16 Konstantin Dobler , Desmond Elliott , Gerard de Melo

Knowledge distillation can lead to deploy-friendly networks against the plagued computational complexity problem, but previous methods neglect the feature hierarchy in detectors. Motivated by this, we propose a general framework for…

Computer Vision and Pattern Recognition · Computer Science 2021-05-13 Yangyang Qin , Hefei Ling , Zhenghai He , Yuxuan Shi , Lei Wu

Knowledge distillation introduced in the deep learning context is a method to transfer knowledge from one architecture to another. In particular, when the architectures are identical, this is called self-distillation. The idea is to feed in…

Machine Learning · Computer Science 2020-10-27 Hossein Mobahi , Mehrdad Farajtabar , Peter L. Bartlett

Dataset distillation aims to synthesize a compact proxy dataset that is unreadable or non-raw from the original dataset for privacy protection and highly efficient learning. However, previous approaches typically adopt a single-stage…

Computer Vision and Pattern Recognition · Computer Science 2026-05-26 Qianxin Xia , Zhiyong Shu , Wenbo Jiang , Jiawei Du , Jielei Wang , Guoming Lu

Object detection has advanced significantly with Detection Transformers (DETRs). However, these models are computationally demanding, posing challenges for deployment in resource-constrained environments (e.g., self-driving cars). Knowledge…

Computer Vision and Pattern Recognition · Computer Science 2025-02-18 Qizhen Lan , Qing Tian

ASR systems often struggle with maintaining syntactic and semantic accuracy in long audio transcripts, impacting tasks like Named Entity Recognition (NER), capitalization, and punctuation. We propose a novel approach that enhances ASR by…

Computation and Language · Computer Science 2025-08-20 Duygu Altinok

Large Language Models (LLMs) are now capable of resolving real-world GitHub issues. However, current approaches overapproximate the code context and suffer from two compounding problems: the prohibitive cost of processing massive inputs,…

Software Engineering · Computer Science 2026-03-31 Haoxiang Jia , Earl T. Barr , Sergey Mechtaev

Knowledge distillation has emerged as a powerful technique for compressing large language models (LLMs) into efficient, deployable architectures while preserving their advanced capabilities. Recent advances in low-rank knowledge…

Machine Learning · Statistics 2026-03-25 Alberlucia Rafael Soarez , Daniel Kim , Mariana Costa , Alejandro Torre

We generalize deep self-attention distillation in MiniLM (Wang et al., 2020) by only using self-attention relation distillation for task-agnostic compression of pretrained Transformers. In particular, we define multi-head self-attention…

Computation and Language · Computer Science 2021-06-29 Wenhui Wang , Hangbo Bao , Shaohan Huang , Li Dong , Furu Wei

Large language models have led to significant progress across many NLP tasks, although their massive sizes often incur substantial computational costs. Distillation has become a common practice to compress these large and highly capable…

Computation and Language · Computer Science 2026-01-06 Zishun Yu , Shangzhe Li , Xinhua Zhang

Traditional approaches to RL have focused on learning decision policies directly from episodic decisions, while slowly and implicitly learning the semantics of compositional representations needed for generalization. While some approaches…

Computation and Language · Computer Science 2022-12-23 Chris Lengerich , Gabriel Synnaeve , Amy Zhang , Hugh Leather , Kurt Shuster , François Charton , Charysse Redwood

This paper discusses our proposal and implementation of Distill, a domain-specific compilation tool based on LLVM to accelerate cognitive models. Cognitive models explain the process of cognitive function and offer a path to human-like…

Programming Languages · Computer Science 2022-01-17 Jan Vesely , Raghavendra Pradyumna Pothukuchi , Ketaki Joshi , Samyak Gupta , Jonathan D. Cohen , Abhishek Bhattacharjee

The remarkable performance of the pre-trained language model (LM) using self-supervised learning has led to a major paradigm shift in the study of natural language processing. In line with these changes, leveraging the performance of speech…

Machine Learning · Computer Science 2021-10-22 Mun-Hak Lee , Joon-Hyuk Chang

In a real-world setting, object instances from new classes can be continuously encountered by object detectors. When existing object detectors are applied to such scenarios, their performance on old classes deteriorates significantly. A few…

Computer Vision and Pattern Recognition · Computer Science 2021-12-16 K J Joseph , Jathushan Rajasegaran , Salman Khan , Fahad Shahbaz Khan , Vineeth N Balasubramanian