Models optimized for real-world tasks reveal the task-dependent necessity of precise temporal coding in hearing
Mark R Saddler, Josh H McDermott · 2024
Read Models optimized for real-world tasks reveal the task-dependent necessity of precise temporal coding in hearing by Mark R Saddler, Josh H McDermott (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-11618365.txt.
Description
Mark Saddler and Josh McDermott present a 2024 study on how precisely the hearing nerve must time its signals. Nerves carry sound partly in the timing of their spikes, and the auditory nerve locks to sound waves with sub-millisecond precision. Whether behavior truly needs that precision has been unclear.
The authors trained machine learning models on simulated ear input to do real hearing tasks. Models with fine timing behaved more like humans at locating sounds and understanding speech. Models without it fell short on those tasks. The amount of precision needed varied by task, though. Some tasks gained from exact timing while others did not. The work shows how training models on real-world jobs can reveal which neural codes perception actually uses.
Catalog record
| CREATOR | Mark R Saddler, Josh H McDermott |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 157 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |