DataClaw0: Agentic Tailoring Multimodal Data from Raw Streams
Summary
DataClaw0 proposes an agentic data tailoring paradigm that uses learnable data processing to structure high-entropy multimodal streams, achieving robust alignment via SFT and GRPO on a novel benchmark.
View Cached Full Text
Cached at: 06/23/26, 09:41 AM
Paper page - DataClaw0: Agentic Tailoring Multimodal Data from Raw Streams
Source: https://huggingface.co/papers/2606.21337 Published on Jun 19
·
Submitted byhttps://huggingface.co/Coneonewan
Congon Jun 23
Abstract
Agentic Data Tailoring paradigm uses learnable data processing to structure high-entropy multimodal streams, with DataClaw_0-9B model achieving robust alignment through SFT and GRPO on a novel benchmark.
Massive unstructuredmultimodal streamssuffer from high “data entropy,” impeding both efficient human knowledge acquisition and high-quality AI post-training. Existing passive annotation paradigms, heavily reliant on heuristic rules or general VLMs, are costly, monotonous, and fail to unlock the deep procedural logic embedded in raw data. We elevate data processing to a learnable capability, proposing a paradigm shift towardsAgentic Data Tailoring, which actively refining and structuring data to align with diverse user and downstream intents. To overcome the data scarcity bottleneck in training such high-order capabilities, we design a two-stage pipeline groundinggenerative semantic synthesisindeterministic Factual Anchors, yielding a large-scale dataset spanning five core physical and digital domains. Building upon this, DataClaw_0-9B model synergizesSupervised Fine-Tuning(SFT) withGroup Relative Policy Optimization(GRPO), achieving robust alignment with complex refinement and tailoring intents. To systematically quantify this capability, we construct DataClaw_0-val, the first benchmark dedicated todata refinement. Crucially, we adoptdownstream post-trainingas the ultimate validation touchstone. Evaluations on video generation, real-world VQA, and GUI navigation confirm that DataClaw_0 delivers high-information-density tailored data, facilitating efficient model adaptation to new tasks under limited training data regimes. Project page: https://czjdsg.github.io/MakeAnyData
View arXiv pageView PDFProject pageGitHub4Add to collection
Get this paper in your agent:
hf papers read 2606\.21337
Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash
Models citing this paper0
No model linking this paper
Cite arxiv.org/abs/2606.21337 in a model README.md to link it from this page.
Datasets citing this paper0
No dataset linking this paper
Cite arxiv.org/abs/2606.21337 in a dataset README.md to link it from this page.
Spaces citing this paper0
No Space linking this paper
Cite arxiv.org/abs/2606.21337 in a Space README.md to link it from this page.
Collections including this paper0
No Collection including this paper
Add this paper to acollectionto link it from this page.
Similar Articles
DataArc-SynData-Toolkit: A Unified Closed-Loop Framework for Multi-Path, Multimodal, and Multilingual Data Synthesis
The article introduces DataArc-SynData-Toolkit, an open-source framework designed to simplify multi-path, multimodal, and multilingual synthetic data generation. It aims to lower technical barriers and improve usability for training large language models through a unified, configuration-driven pipeline.
VisualClaw: A Real-Time, Personalized Agent for the Physical World
VisualClaw is a self-evolving multimodal agent that reduces deployment costs through hybrid encoding and skill evolution, while improving video-QA accuracy across multiple benchmarks.
CL-DMDF:Dynamic Multimodal Data Fusion Model Based on Contrastive Learning
This paper proposes CL-DMDF, a dynamic multimodal data fusion model that uses contrastive learning and a dual-dimensional attention mechanism to handle missing modalities and improve discriminative learning.
Alignment Tuning for Large Language Models: A Data-Centric Lens on Alignment Data Pipelines
This survey reframes the alignment tuning of large language models as a data pipeline design problem, decomposing it into three stages: response synthesis, preference evaluation, and preference instantiation. It identifies design trade-offs and failure modes, and outlines open challenges such as prompt-level alignment and agentic settings.
Constraint-Aware Flow Matching: Decision Aligned End-to-End Training for Constrained Sampling
Proposes Constraint-Aware Flow Matching, a novel end-to-end framework that aligns the model's learning dynamics with constrained sampling procedure, mitigating distributional shift from projection corrections for high-quality constrained generation.