DataClaw0: Agentic Tailoring Multimodal Data from Raw Streams
For practitioners in multimodal AI post-training, this work addresses the bottleneck of costly and monotonous data annotation by proposing a learnable data tailoring capability, though the approach is incremental as it builds on existing SFT and GRPO methods.
DataClaw0 introduces a paradigm shift from passive annotation to active agentic data tailoring, enabling efficient refinement of raw multimodal streams. The 9B model achieves robust alignment with complex intents, delivering high-information-density data that facilitates efficient model adaptation under limited training data, as validated on video generation, VQA, and GUI navigation tasks.
Massive unstructured multimodal streams suffer 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 towards Agentic 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 grounding generative semantic synthesis in deterministic Factual Anchors, yielding a large-scale dataset spanning five core physical and digital domains. Building upon this, $\text{DataClaw}_0$-9B model synergizes Supervised Fine-Tuning (SFT) with Group Relative Policy Optimization (GRPO), achieving robust alignment with complex refinement and tailoring intents. To systematically quantify this capability, we construct $\text{DataClaw}_0$-val, the first benchmark dedicated to data refinement. Crucially, we adopt downstream post-training as the ultimate validation touchstone. Evaluations on video generation, real-world VQA, and GUI navigation confirm that $\text{DataClaw}_0$ 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