Test-Time Instance-Specific Parameter Composition: A New Paradigm for Adaptive Generative Modeling

• URL: http://arxiv.org/abs/2603.27665v1
• Date: Sun, 29 Mar 2026 12:31:33 GMT
• Title: Test-Time Instance-Specific Parameter Composition: A New Paradigm for Adaptive Generative Modeling
• Abstract summary: Composer is a new paradigm for adaptive generative modeling based on test-time instance-specific parameter composition.<n>By leveraging input-aware parameter composition, Composer establishes a new paradigm for designing generative models that dynamically adapt to each input.
• Score: 38.42229422362911
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Existing generative models, such as diffusion and auto-regressive networks, are inherently static, relying on a fixed set of pretrained parameters to handle all inputs. In contrast, humans flexibly adapt their internal generative representations to each perceptual or imaginative context. Inspired by this capability, we introduce Composer, a new paradigm for adaptive generative modeling based on test-time instance-specific parameter composition. Composer generates input-conditioned parameter adaptations at inference time, which are injected into the pretrained model's weights, enabling per-input specialization without fine-tuning or retraining. Adaptation occurs once prior to multi-step generation, yielding higher-quality, context-aware outputs with minimal computational and memory overhead. Experiments show that Composer substantially improves performance across diverse generative models and use cases, including lightweight/quantized models and test-time scaling. By leveraging input-aware parameter composition, Composer establishes a new paradigm for designing generative models that dynamically adapt to each input, moving beyond static parameterization.

Related papers

• Constraint-based Pre-training: From Structured Constraints to Scalable Model Initialization [29.901255886542515]
  WeiT is a pre-training paradigm that disentangles size-agnostic knowledge into reusable weight templates.<n>We show that WeiT achieves state-of-the-art performance in initializing models with varying depths and widths.
  arXiv  Detail & Related papers  (2026-04-16T08:29:48Z)
• HY-WU (Part I): An Extensible Functional Neural Memory Framework and An Instantiation in Text-Guided Image Editing [0.0]
  We propose HY-WU (Weight Unleashing), a memory-first adaptation framework.<n>It shifts adaptation pressure away from overwriting a single shared parameter point.<n> HY-WU implements functional (operator-level) memory as a neural module.
  arXiv  Detail & Related papers  (2026-03-07T14:40:05Z)
• Continual Adaptation: Environment-Conditional Parameter Generation for Object Detection in Dynamic Scenarios [54.58186816693791]
  environments constantly change over time and space, posing significant challenges for object detectors trained based on a closed-set assumption.<n>We propose a new mechanism, converting the fine-tuning process to a specific- parameter generation.<n>In particular, we first design a dual-path LoRA-based domain-aware adapter that disentangles features into domain-invariant and domain-specific components.
  arXiv  Detail & Related papers  (2025-06-30T17:14:12Z)
• Merging Models on the Fly Without Retraining: A Sequential Approach to Scalable Continual Model Merging [75.93960998357812]
  Deep model merging represents an emerging research direction that combines multiple fine-tuned models to harness their capabilities across different tasks and domains.<n>Current model merging techniques focus on merging all available models simultaneously, with weight matrices-based methods being the predominant approaches.<n>We propose a training-free projection-based continual merging method that processes models sequentially.
  arXiv  Detail & Related papers  (2025-01-16T13:17:24Z)
• Optimize Incompatible Parameters through Compatibility-aware Knowledge Integration [104.52015641099828]
  Existing research excels in removing such parameters or merging the outputs of multiple different pretrained models.<n>We propose Compatibility-aware Knowledge Integration (CKI), which consists of Deep Assessment and Deep Splicing.<n>The integrated model can be used directly for inference or for further fine-tuning.
  arXiv  Detail & Related papers  (2025-01-10T01:42:43Z)
• Superposition in Transformers: A Novel Way of Building Mixture of Experts [0.0]
  Catastrophic forgetting is a major challenge when adapting large language models to new tasks or domains.<n>We introduce Superposition in Transformers, a novel architecture that leverages autoencoders to superimpose the hidden representations of a base model.
  arXiv  Detail & Related papers  (2024-12-31T16:28:23Z)
• Parameter Competition Balancing for Model Merging [13.66727853299506]
  PCB-Merging is a training-free technique that adjusts the coefficients of each parameter for effective model merging. PCB-Merging achieves substantial performance enhancements across multiple modalities, domains, model sizes, number of tasks, fine-tuning forms, and large language models.
  arXiv  Detail & Related papers  (2024-10-03T11:17:58Z)
• Majority Kernels: An Approach to Leverage Big Model Dynamics for Efficient Small Model Training [32.154166415680066]
  Methods like distillation, compression or quantization help leverage the highly performant large models to induce smaller performant ones. This paper explores the hypothesis that a single training run can simultaneously train a larger model for performance and derive a smaller model for deployment.
  arXiv  Detail & Related papers  (2024-02-07T17:07:41Z)
• Exploring and Evaluating Personalized Models for Code Generation [9.25440316608194]
  We evaluate transformer model fine-tuning for personalization. We consider three key approaches: (i) custom fine-tuning, which allows all the model parameters to be tuned. We compare these fine-tuning strategies for code generation and discuss the potential generalization and cost benefits of each in various deployment scenarios.
  arXiv  Detail & Related papers  (2022-08-29T23:28:46Z)
• HyperImpute: Generalized Iterative Imputation with Automatic Model Selection [77.86861638371926]
  We propose a generalized iterative imputation framework for adaptively and automatically configuring column-wise models. We provide a concrete implementation with out-of-the-box learners, simulators, and interfaces.
  arXiv  Detail & Related papers  (2022-06-15T19:10:35Z)
• Re-parameterizing Your Optimizers rather than Architectures [119.08740698936633]
  We propose a novel paradigm of incorporating model-specific prior knowledge into Structurals and using them to train generic (simple) models. As an implementation, we propose a novel methodology to add prior knowledge by modifying the gradients according to a set of model-specific hyper- parameters. For a simple model trained with a Repr, we focus on a VGG-style plain model and showcase that such a simple model trained with a Repr, which is referred to as Rep-VGG, performs on par with the recent well-designed models.
  arXiv  Detail & Related papers  (2022-05-30T16:55:59Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.