Capability Provenance in Language Models: A Case Study in Social Reasoning
We use training-data attribution as an interpretable tool for capability discovery, mapping which regions of the pretraining corpus support social reasoning versus STEM reasoning in OLMo3-7B. Training-data attribution measures how strongly each training document influences a model's predictions on a benchmark, but docu...
Glenn Scott Matlin, Chandreyi Chakraborty, Saehee Eom et al.
· 0 citations