AI Model Modulation with Logits Redistribution
Large-scale models are typically adapted to meet the diverse requirements of model owners and users. However, maintaining multiple specialized versions of the model is inefficient. In response, we propose AIM, a novel model modulation paradigm that enables a single model to exhibit diverse behaviors to meet the specific end requirements. AIM enables two key modulation modes: utility and focus modulations. The former provides model owners with dynamic control over output quality to deliver varying utility levels, and the latter offers users precise control to shift model's focused input features. AIM introduces a logits redistribution strategy that operates in a training data-agnostic and retraining-free manner. We establish a formal foundation to ensure AIM's regulation capability, based on the statistical properties of logits ordering via joint probability distributions. Our evaluation confirms AIM's practicality and versatility for Al model modulation, with tasks spanning image classification, semantic segmentation and text generation, and prevalent architectures including ResNet, SegFormer and Llama.
Code (0)
등록된 구현이 없습니다.
Tasks
Semantic SegmentationImage ClassificationText GenerationSimilar Papers 제목 키워드 기반
Direct Segmentation without Logits Optimization for Training-Free Open-Vocabulary Semantic Segmentation
Open-vocabulary semantic segmentation (OVSS) aims to segment arbitrary category regions in images using open-vocabulary prompts, necessitating that existing methods possess pixel-level vision-language alignment capabilit…
Semantic SegmentationRedistribution Systems and PRAM
Redistribution systems iteratively redistribute mass between groups under the control of rules. PRAM is a framework for building redistribution systems. We discuss the relationships between redistribution systems, agent-…
A Cognitive Architecture for the Implementation of Emotions in Computing Systems
In this paper we present a new neurobiologically-inspired affective cognitive architecture: NEUCOGAR (NEUromodulating COGnitive ARchitecture). The objective of NEUCOGAR is the identification of a mapping from the influen…
Redistribution Mechanism on Networks
Redistribution mechanisms have been proposed for more efficient resource allocation but not for profit. We consider redistribution mechanism design in a setting where participants are connected and the resource owner is …
Redistribution with Needs
We take an axiomatic approach to study redistribution problems when agents report income and needs. We formalize axioms reflecting ethical and operational principles such as additivity, impartiality and individual ration…