Realmbird/

nla-thought-anchors-hash_activations-step1

GSM8K test-set activations and NLA verbalisations from Qwen2.5-7B-Instruct, extracted at layer 20 at the #### token of the model's generated response. Part of the nla-thought-anchors project, replicating and extending Ryan Greenblatt's findings on Natural Language Autoencoders.

	Files




	

File Description

correct.parquet Examples where model answer matched gold (708 rows)

incorrect.parquet Examples where model answer did not match… See the full description on the dataset page: https://huggingface.co/datasets/Realmbird/nla-thought-anchors-hash_activations-step1.