Skip to content
conftrace_

pretraining data mixture

2 papers

Explore in graph

Co-occurring keywords · papers mentioning both

label-shift assumption (1) constrained inverse problem (1) confusion matrix calibration (1) post-hoc auditing (1) m2-bert (1) monarch mixer (1) single-sample batches (1) state-space encoder (1) data mixture surgery (1) long-context retrieval (1)

Papers

LLMSurgeon: Diagnosing Data Mixture of Large Language Models ACL 2026 Benchmarking and Building Long-Context Retrieval Models with LoCo and M2-BERT ICML 2024