DRIFT: A Residual Flow Adapter for Decoding Continuous Outputs in Vision-Language Models

• URL: http://arxiv.org/abs/2606.05758v1
• Date: Thu, 04 Jun 2026 06:37:10 GMT
• Title: DRIFT: A Residual Flow Adapter for Decoding Continuous Outputs in Vision-Language Models
• Abstract summary: We propose DRIFT, a framework for adapting pretrained vision-language models to continuous decoding tasks.<n>DRIFT combines a base predictor, which provides a coarse estimate of the target output, with a generative refinement module based on flow matching.<n>We evaluate DRIFT on both perception and planning tasks, including visual grounding and robotic control.
• Score: 9.25715767290436
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Many modern vision-language models (VLMs) build on autoregressive decoding of discrete tokens. While text-based output interfaces enable scalable pretraining and strong zero-shot generalization across diverse tasks, they are poorly suited for problems that require precise continuous outputs, such as localizing temporal boundaries of events or generating robotic control actions. To address this challenge, we propose DRIFT, a general framework for adapting pretrained VLMs to continuous decoding tasks. DRIFT combines a base predictor, which provides a coarse estimate of the target output, with a generative refinement module based on flow matching that iteratively improves the prediction. This residual formulation transforms the generative modeling problem from learning a global output distribution to modeling a localized residual distribution around a strong prior, substantially simplifying optimization. We evaluate DRIFT on both perception and planning tasks, including visual grounding and robotic control. Across multiple tasks and architectures spanning MLLMs, VLAs, and WAMs, DRIFT consistently outperforms a strong set of regression- and generative-based solutions.

Related papers

• Diffusion Language Models for Mobile Edge Agentic AI: Foundations, Applications, and Challenges [58.887152635127165]
  Diffusion language models (DLMs) offer a non-autoregressive alternative for mobile edge agentic artificial intelligence (AI)<n>DLMs can update multiple uncertain tokens in parallel and exploit bidirectional context throughout the generation process.<n>This survey reviews DLM foundations and analyzes their suitability for edge settings under latency, memory, energy, bandwidth, privacy, and reliability constraints.
  arXiv  Detail & Related papers  (2026-09-04T06:16:49Z)
• Continual Learning with Elastic Regularization and Synthetic Replay for Federated MLLM Fine-Tuning [15.445619998280243]
  Federated Continual Multimodal Learning (FedCMM) is a framework that embeds continual-learning safeguards into the federated optimization loop.<n>FedCMM consistently outperforms recent baselines on accuracy and backward transfer.
  arXiv  Detail & Related papers  (2026-07-13T19:40:30Z)
• Modular Foundation Models for Time-Series Perception in Digital Twins [0.8983000676504388]
  Engineering Digital Twins and Prognostics and Health Management (PHM) systems rely on robust perception modules to extract actionable information from time-series data.<n>This paper proposes a modular foundation model for time-series perception based on a collection of pretrained representation encoders.<n>The framework leverages self-supervised learning on heterogeneous datasets to learn transferable and task-agnostic representations, which can be reused across multiple PHM tasks.
  arXiv  Detail & Related papers  (2026-07-03T20:02:28Z)
• FLARE: Diffusion for Hybrid Language Model [72.60770374799634]
  FLARE is a systematic conversion framework for hybrid-attention large language models.<n>It combines a token-equal AR-and-diffusion objective, hardware-aware kernels, and unified inference, enabling one checkpoint to support both AR-style verified decoding and diffusion-style parallel denoising.<n>Our results suggest practical dLLMs are limited not only by decoding algorithms, but also by transfer data quality and the training inefficiency of current block-diffusion objectives.
  arXiv  Detail & Related papers  (2026-06-01T06:58:15Z)
• Optimization-as-a-Service via Multi-Agent Large Language Model for Radio Access Networks [33.1158889975007]
  Sixth-generation (6G) RAN environments confront unprecedented service diversity and exponential dynamics.<n>Manual models and standard AI algorithms remain fundamentally rigid, lacking the flexibility to adapt and self-evolve.<n>We propose treating the PRB allocation problem as an optimization reflection-as-a-Service (O) system.<n>This fundamentally reshapes RAN resource allocation by utilizing agents to dynamically construct optimization problems and automatically determine objectives tailored to real-time scenarios.
  arXiv  Detail & Related papers  (2026-05-15T14:14:29Z)
• Efficient and Adaptive Human Activity Recognition via LLM Backbones [39.42078885809324]
  We propose a paradigm shift that reuses large pretrained language models (LLMs) as generic temporal backbones for sensor-based HAR.<n>We show that this approach enables rapid convergence, strong data efficiency, and robust cross-dataset transfer.
  arXiv  Detail & Related papers  (2026-05-12T12:06:39Z)
• OneDrive: Unified Multi-Paradigm Driving with Vision-Language-Action Models [69.2503510410147]
  We present a unified autonomous driving framework built upon a pretrained VLM.<n>We demonstrate that pretrained VLM attention exhibits strong transferability beyond pure language modeling.<n>Experiments on end-to-end autonomous driving benchmarks demonstrate state-of-the-art performance.
  arXiv  Detail & Related papers  (2026-04-20T07:50:00Z)
• Reshaping Action Error Distributions for Reliable Vision-Language-Action Models [69.38615670891038]
  In robotic manipulation, vision-language-action (VLA) models have emerged as a promising paradigm for learning generalizable and scalable robot policies.<n>We focus on continuous-action VLA models and move beyond conventional MSE-based regression by reshaping action error distributions during training.<n>We evaluate our approaches across standard, few-shot, and noisy settings on multiple representative VLA architectures.
  arXiv  Detail & Related papers  (2026-02-04T05:37:09Z)
• LP-LLM: End-to-End Real-World Degraded License Plate Text Recognition via Large Multimodal Models [4.497411606350301]
  Real-world License Plate Recognition (LPR) faces significant challenges from severe degradations such as motion blur, low resolution, and complex illumination.<n>The prevailing "restoration-then-recognition" two-stage paradigm suffers from a fundamental flaw: the pixel-level optimization objectives of image restoration models are misaligned with the semantic goals of character recognition.<n>We propose an end-to-end structure-aware multimodal reasoning framework based on Qwen3-VL.
  arXiv  Detail & Related papers  (2026-01-14T03:32:55Z)
• Co-Training Vision Language Models for Remote Sensing Multi-task Learning [68.15604397741753]
  Vision language models (VLMs) have achieved promising results in RS image understanding, grounding, and ultra-high-resolution (UHR) image reasoning.<n>We present RSCoVLM, a simple yet flexible VLM baseline for RS MTL.<n>We propose a unified dynamic-resolution strategy to address the diverse image scales inherent in RS imagery.
  arXiv  Detail & Related papers  (2025-11-26T10:55:07Z)
• Scaling LLM Speculative Decoding: Non-Autoregressive Forecasting in Large-Batch Scenarios [76.85739138203014]
  We present SpecFormer, a novel architecture that accelerates unidirectional and attention mechanisms.<n>We demonstrate that SpecFormer achieves lower training demands and reduced computational costs.
  arXiv  Detail & Related papers  (2025-11-25T14:20:08Z)
• ImagiDrive: A Unified Imagination-and-Planning Framework for Autonomous Driving [64.12414815634847]
  Vision-Language Models (VLMs) and Driving World Models (DWMs) have independently emerged as powerful recipes addressing different aspects of this challenge.<n>We propose ImagiDrive, a novel end-to-end autonomous driving framework that integrates a VLM-based driving agent with a DWM-based scene imaginer.
  arXiv  Detail & Related papers  (2025-08-15T12:06:55Z)
• RCCDA: Adaptive Model Updates in the Presence of Concept Drift under a Constrained Resource Budget [28.53294084812961]
  Real-time machine learning algorithms are often faced with the challenge of adapting models to concept drift.<n>Existing solutions often depend on drift-detection methods that produce high computational overhead for resource-constrained environments.<n>We propose RCCDA: a dynamic model update policy that optimize ML training dynamics while ensuring compliance to predefined resource constraints.
  arXiv  Detail & Related papers  (2025-05-30T02:49:42Z)

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.