PODS is a plug-and-play oscillatory data-volume scheduler that alternates low-ratio regularization phases with high-ratio recovery phases to improve data selection efficiency across training tasks.
D2 pruning: Message passing for balancing diversity and difficulty in data pruning
6 Pith papers cite this work. Polarity classification is still indexing.
6
Pith papers citing it
years
2026 6verdicts
UNVERDICTED 6representative citing papers
Proposes memorization-guided two-stage scoring to select debiased training subsets, enabling ERM models to achieve better performance than SOTA debiasing techniques using only 10% of data.
MAGIC selects coreset subsets for VLM instruction tuning via Multimodal Gain, Bridging Relevance, and Skill-Neuron Signatures, reporting 100.3% and 101.6% relative performance to full finetuning at 20% budget on LLaVA-665K and Vision-Flan with 73.7% less runtime.
A video transfer pipeline augments simulated VLA data into realistic videos while preserving actions, yielding consistent performance gains on robot benchmarks such as 8% on Robotwin 2.0.
POES frames prompt evaluation as online adaptive testing and uses a provably submodular objective to pick informative examples, delivering 6.2% higher average accuracy and 35-60% token savings versus naive full-set scoring.
Off-the-shelf models assess quality and alignment to select diverse multimodal training data, letting models trained on the filtered subset match or exceed full-dataset results on standard benchmarks.
citing papers explorer
-
Beyond What to Select: A Plug-and-play Oscillatory Data-Volume Scheduling for Efficient Model Training
PODS is a plug-and-play oscillatory data-volume scheduler that alternates low-ratio regularization phases with high-ratio recovery phases to improve data selection efficiency across training tasks.
-
Mitigating Spurious Correlations with Memorization-Guided Dataset De-Biasing
Proposes memorization-guided two-stage scoring to select debiased training subsets, enabling ERM models to achieve better performance than SOTA debiasing techniques using only 10% of data.
-
MAGIC: Multimodal Alignment & Grounding-aware Instruction Coreset for Vision-Language Models
MAGIC selects coreset subsets for VLM instruction tuning via Multimodal Gain, Bridging Relevance, and Skill-Neuron Signatures, reporting 100.3% and 101.6% relative performance to full finetuning at 20% budget on LLaVA-665K and Vision-Flan with 73.7% less runtime.
-
Seeing Realism from Simulation: Efficient Video Transfer for Vision-Language-Action Data Augmentation
A video transfer pipeline augments simulated VLA data into realistic videos while preserving actions, yielding consistent performance gains on robot benchmarks such as 8% on Robotwin 2.0.
-
Select Smarter, Not More: Prompt-Aware Evaluation Scheduling with Submodular Guarantees
POES frames prompt evaluation as online adaptive testing and uses a provably submodular objective to pick informative examples, delivering 6.2% higher average accuracy and 35-60% token savings versus naive full-set scoring.
-
DOSE: Data Selection for Multi-Modal LLMs via Off-the-Shelf Models
Off-the-shelf models assess quality and alignment to select diverse multimodal training data, letting models trained on the filtered subset match or exceed full-dataset results on standard benchmarks.