7 papers
HERMES: A Multi-Granularity Labeling Substrate for Pre-training Data Mixtures
Ziyun Qiao, Yue Min, Ruining Chen +1
Most data-mixing methods assume the corpus has already been partitioned into groups, and the choice of those groups determines what a mixer can express. Existing labels, including…
GRASP: Geometry-aware Residual Alignment for Scalable Pretraining Data Attribution
Yue Min, Ruining Chen, Yujun Li
Scalable data attribution methods typically assign isolated utility scores to individual training examples. This prevalent additive assumption fundamentally fails to capture critic…
GEM: Geometric Entropy Mixing for Optimal LLM Data Curation
Yue Min, Ziyun Qiao, Ruining Chen +1
LLM pre-training efficacy increasingly depends on data composition rather than sheer volume. Yet, optimal mixing is hindered by categorization flaws: human taxonomies suffer from o…
Winning the Pruning Gamble: A Unified Approach to Joint Sample and Token Pruning for Efficient Supervised Fine-Tuning
Shaobo Wang, Jiaming Wang, Jiajun Zhang +9
As supervised fine-tuning (SFT) evolves from a lightweight post-training step into a compute-intensive phase rivaling mid-training in scale, data efficiency has become critical for…
ImagebindDC: Compressing Multi-modal Data with Imagebind-based Condensation
Yue Min, Shaobo Wang, Jiaze Li +5
Data condensation techniques aim to synthesize a compact dataset from a larger one to enable efficient model training, yet while successful in unimodal settings, they often fail in…
Rethinking LLM Evaluation: Can We Evaluate LLMs with 200x Less Data?
Shaobo Wang, Cong Wang, Wenjie Fu +11
As the demand for comprehensive evaluations of diverse model capabilities steadily increases, benchmark suites have correspondingly grown significantly in scale. Despite notable ad…