specific method · filed under data curation
Recycling filtered documents into image-text pairs
Some filtered documents are screened and recombined to produce additional image-text pairs.
- source
- 1
- model
- 1
- lab adopt it
- 1
- strongest
- used
How sources treat it
One count per evidence span, weakest treatment to strongest.
used 2
Documented in
Evidence
2 spans quoted from the sources, strongest treatment first.
Documents filtered out during this process are partly recycled into additional image-text pairs via screening and recombination.
usedunclearin DeepSeek-V4.1-FlashDeepSeek
Documents filtered out during this process are partly recycled into additional image-text pairs via screening and recombination
usedunclearin DeepSeek-V4.1-FlashDeepSeek
Filed alongside
Other methods under data curation :: synthetic data.
Agentic data synthesis pipelineKnowledge distillation for synthetic dataAutomated data synthesis and environment-construction pipelinesAutomatic synthesis of task-oriented RL environmentsContainerized coding-environment construction with self-testing and trace removalCounterfactual data augmentationAutomated batch synthesis of RL training dataCoding environments from GitHub issue–PR pairsContainer buildability and verifiability checkCorpus rephrasing with fidelity verificationDynamic multi-agent synthetic-data generation loopEnd-to-end synthetic environment generationEnvironment, toolset, task, and solution synthesis pipelineFailure-case and negative-feedback-driven environment generationHeterogeneous answer-generation agentsIterative task-difficulty escalationKnowledge-graph-guided task synthesisLarge-scale environment synthesis and curationLong-context data synthesis by permutation and concatenationMatching synthesis-pipeline complexity to teacher capabilityMetadata-conditioned synthetic generationMocked tools for agent environmentsModular synthetic-data pipeline compositionMulti-agent collaborative environment construction