Contributed Talk Sessions | Poster Sessions | All Posters | Search Papers

Poster Session A: Tuesday, August 12, 1:30 – 4:30 pm, de Brug & E‑Hall

Introducing the CORTEX Database: COntext-dependent Reinforcement learning and Transfer EXperiments

Isabelle Hoxha¹, Stefano Palminteri²; ¹Leiden University, ²Ecole Normale Supérieure – PSL

Presenter: Isabelle Hoxha

Learning - Transfer paradigms are particularly relevant to study value learning, as they can uncover contextual value learning. However, there is no consensus within and across laboratories on the implementation of the experimental variable nor the exact computational processing underlying context-dependence, thus making it unclear whether models such as reference-point and range adaptation genuinely generalize across contexts or are merely tailored to specific task structures. We therefore created an extremely large (n>2500 human participants) yet extremely curated dataset to establish a standardized framework and systematically evaluate the applicability of competing value learning models. We showed that the range adaptation model is the best fitting model for half of the participants, with task specifications such as feedback type modulating the relevance of in-context learning models.

Topic Area: Memory, Spatial Cognition & Skill Learning

Extended Abstract: Full Text PDF