The temporal dynamics of animacy-related neural representations are robust to controversial training

In 30 healthy participants, feedback-based animacy categorisation training did not produce a statistically significant reduction of EEG animal bias for lookalikes, despite improved categorisation efficiency.

• Why it matters: Animal bias is robust across domains, but it is unclear whether learning with explicit feedback can reshape visual neural representations toward inanimate interpretations.
• What they did: Participants completed three weekly sessions of feedback-based training to categorise lookalikes as inanimate, then performed an EEG animacy task; behavioural inverse efficiency improved, and EEG time-resolved multivariate pattern analysis plus representational similarity analysis were compared to a no-training dataset.
• The result: EEG analyses showed only a descriptively smaller animal bias after training, while a significant animal bias remained, implying limited neural reorganisation after several sessions.

AI summary of the abstract · may contain errors · not medical advice · read the paper.