MIRA discovers source-specific rubrics then distills them into scalable scorers, matching full-corpus mid-training performance on code tasks while using half the tokens.
Prop.” is the per-source sampling proportion applied at stage-3 over the stage-2 low / mid / high bins. “Rows
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
MIRA: Mid-training Rubric Anchoring for Source-Aware Data Selection
MIRA discovers source-specific rubrics then distills them into scalable scorers, matching full-corpus mid-training performance on code tasks while using half the tokens.