Successor representation supports structural learning of syllable sequences
- 1Department of Neurology, Duke University, Durham, NC 27705, USA
Abstract
Humans efficiently learn the temporal structure of speech, yet the underlying cognitive mechanisms remain unclear. Recent research in visuospatial sequential memory has proposed the successor representation (SR), important in reinforcement learning, which encodes multi-step transitional relationships. To test whether SR also supports the structural learning of rapidly changing speech syllable sequences, we adapted a pyramid-structured sequential learning paradigm using syllable streams presented at a natural speech rate (~ 4 Hz). Participants exhibited robust structural learning, evidenced by asymmetric memory performance modulated by adjacent transitional probabilities. Notably, the SR model captured recognition performance for non-adjacent syllable pairs. Moreover, SR-based learning effects generalized across all syllable sequences, regardless of adjacency. Furthermore, the effectiveness of SR learning was associated with cognitive flexibility and working memory capacity. Together, these findings suggest that humans rely on effortful SR-like mechanisms for the structural learning of syllable sequences, highlighting a modality-general computational principle for sequential learning across various temporal modulation scales.
Related
- High-dimensional hippocampus representation facilitates learning of low-dimensional abstract representation and supports generalization
- A representation of the visual environment in the human basolateral amygdala supports learning and choice
- HyCoSeq: Contextual Hyperbolic Representation Learning for Genomic Sequences
- Multimodal Protein Retrieval via Joint Representation Learning from Sequences and Cryo-EM Density Maps
- How thermostable direct haemolysin (TDH) diverges from TDH-related haemolysin (TRH)? Reassessing haemolysins in Vibrio parahaemolyticus through functional and structural representation learning
BibTeX
@misc{liu2026successor,
title = {{Successor representation supports structural learning of syllable sequences}},
author = {J. Liu and Y. Wang and J. Feng and H. Zhao and Q. He and X. Yang and B. Liang},
howpublished = {bioRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.03.756368},
url = {https://www.biorxiv.org/content/10.64898/2026.10.03.756368v1},
note = {CC BY-NC-ND},
}
RIS
TY - UNPB
TI - Successor representation supports structural learning of syllable sequences
AU - J. Liu
AU - Y. Wang
AU - J. Feng
AU - H. Zhao
AU - Q. He
AU - X. Yang
AU - B. Liang
PY - 2026
DA - 2026/10/03
PB - bioRxiv
DO - 10.64898/2026.10.03.756368
UR - https://www.biorxiv.org/content/10.64898/2026.10.03.756368v1
AB - Humans efficiently learn the temporal structure of speech, yet the underlying cognitive mechanisms remain unclear. Recent research in visuospatial sequential memory has proposed the successor representation (SR), important in reinforcement learning, which encodes multi-step transitional relationships. To test whether SR also supports the structural learning of rapidly changing speech syllable sequences, we adapted a pyramid-structured sequential learning paradigm using syllable streams presented at a natural speech rate (~ 4 Hz). Participants exhibited robust structural learning, evidenced by asymmetric memory performance modulated by adjacent transitional probabilities. Notably, the SR model captured recognition performance for non-adjacent syllable pairs. Moreover, SR-based learning effects generalized across all syllable sequences, regardless of adjacency. Furthermore, the effectiveness of SR learning was associated with cognitive flexibility and working memory capacity. Together, these findings suggest that humans rely on effortful SR-like mechanisms for the structural learning of syllable sequences, highlighting a modality-general computational principle for sequential learning across various temporal modulation scales.
ER -