SciImpact is a new multi-dimensional benchmark showing that multi-task fine-tuned smaller LLMs can outperform larger models in predicting varied forms of scientific impact.
Ching Jin, Yifang Ma, and Brian Uzzi
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
SciImpact: A Multi-Dimensional, Multi-Field Benchmark for Scientific Impact Prediction
SciImpact is a new multi-dimensional benchmark showing that multi-task fine-tuned smaller LLMs can outperform larger models in predicting varied forms of scientific impact.