paper-with-me

홈 › Papers

Dynamic Activation Pitfalls in LLaMA Models: An Empirical Study

2024-05-15 · Chi Ma, Mincong Huang, Chao Wang, Yujie Wang, Lei Yu

In this work, we systematically investigate the efficacy of dynamic activation mechanisms within the LLaMA family of language models. Despite the potential of dynamic activation methods to reduce computation and increase speed in models using the ReLU activation function, our empirical findings have uncovered several inherent pitfalls in the current dynamic activation schemes. Through extensive experiments across various dynamic activation strategies, we demonstrate that LLaMA models usually underperform when compared to their ReLU counterparts, particularly in scenarios demanding high sparsity ratio. We attribute these deficiencies to a combination of factors: 1) the inherent complexity of dynamically predicting activation heads and neurons; 2) the inadequate sparsity resulting from activation functions; 3) the insufficient preservation of information resulting from KV cache skipping. Our analysis not only sheds light on the limitations of dynamic activation in the context of large-scale LLaMA models but also proposes roadmaps for enhancing the design of future sparsity schemes.

📄 PDF Abstract BibTeX arXiv:2405.09274

Code (0)

등록된 구현이 없습니다.

Tasks

Attribute

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…
LLaMA LLaMA is a collection of foundation language models ranging from 7B to 65B parameters. It is based on the transformer architecture with various improvements that were…

Similar Papers 제목 키워드 기반

The Pitfalls of KV Cache Compression

2025-09-30 · Alex Chen, Renato Geh, Aditya Grover, Guy Van den Broeck 외 arxiv

KV cache compression promises increased throughput and efficiency with negligible loss in performance. While the gains in throughput are indisputable and recent literature has indeed shown minimal degradation on particul…

The Uniqueness of LLaMA3-70B Series with Per-Channel Quantization

2024-08-27 · Minghai Qin

We have observed a distinctive quantization-related behavior in the LLaMA3/3.1-70B models that is absent in both the LLaMA2-70B and LLaMA3/3.1/3.2-1B/3B/8B/405B models. Quantization is a crucial technique for deploying l…

Quantization

MOYU: A Theoretical Study on Massive Over-activation Yielded Uplifts in LLMs

2024-06-18 · Chi Ma, Mincong Huang, Chao Wang, Yujie Wang 외

Massive Over-activation Yielded Uplifts(MOYU) is an inherent property of large language models, and dynamic activation(DA) based on the MOYU property is a clever yet under-explored strategy designed to accelerate inferen…

Identifying and Manipulating Personality Traits in LLMs Through Activation Engineering

2024-12-10 · Rumi A. Allbert, James K. Wiles, Vlad Grankovsky

The field of large language models (LLMs) has grown rapidly in recent years, driven by the desire for better efficiency, interpretability, and safe use. Building on the novel approach of "activation engineering," this st…

ActTail: Global Activation Sparsity in Large Language Models

2026-02-18 · Wenwen Hou, Xinyuan Song, Shiwei Liu arxiv

Activation sparsity is a promising approach for accelerating large language model (LLM) inference by reducing computation and memory movement. However, existing activation sparsity methods typically apply uniform sparsit…