fMTP: A Unifying Computational Framework of Temporal Preparation Across Time Scales

University of Groningen

fMTP

Salet, Josh; Kruijne, Wouter; Rijn, van, Hedderik

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from

it. Please check the document version below.

Publication date:

2019

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):

Salet, J., Kruijne, W., & Rijn, van, H. (2019). fMTP: A Unifying Computational Framework of Temporal

Preparation Across Time Scales.

Copyright

Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policy

If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.

Josh M. Salet

1*

, Wouter Kruijne

1

, Sander A. Los

2

*j.m.salet@rug.nl

1

_{University of Groningen}

2

_{Vrije Universiteit Amsterdam}

Time

_Motor

FP

Time Cells

500

1000

_{1500 t(ms)}

500

1000

1500 t(ms)

The core of fMTP is an implicit associative learning

process between time cells and activation- and

inhibition motor units

Temporal preparation is

modulated across a

vast

range of time scales

Trace Formation

Trace Retrieval

Distribution

Previous trial

Across blocks

This project was partially supported by the research programme "Interval Timing in the Real World: A functional, computational and neuroscience approach", project number 453-16-005, financed by the NWO

We pursue the view

that these phenomena

are driven by

implicitly

learned processes

Hazard accounts: Preparation is controlled by a

process that follows the conditional probability

fMTP integrates

insights on the

coding of

time, episodic memory,

and motor planning

Previous experiences drive preparation according their recency

Strong influence of recent trials

gives rise to short-term dynamics

Long-lasting influence of each trial

gives rise to long-term dynamics

We show the data to be consistent with fMTP and

to deviate from the Hazard function

Previous trial

?

What is the unifying

mechanism giving rise to

these phenomena?

ITI

_S1

FP

S2

+ +

+

+

+

+

fMTP: A Unifying Computational Framework

of Temporal Preparation

Across Time Scales

Hedderik van Rijn

1

, Martijn Meeter

2

Simulations

demonstrate that

fMTP

captures all phenomena

FP Paradigm

Probabili

ty

Foreperiod (ms)

0.50

0.25

400

800

1200

1600

Foreperiod (ms)

400

800

1200

1600

Group Anti-exp

Group Exp

Foreperiod (ms)

400

800

1200

1600

Group Anti-exp

Group Exp

Pre-acquisition

Acquisition

_Test

500 1000 1500

400

375

350

325

500

Data

Model

1000 1500

500 1000 1500

Exp

FP Distribution (ms)

400

800

1200

1600

Anti-exp

Uniform

Probabili

ty

0.50

0.25

Trace Expression

Memory Strength

(a.u.)

Previous Trial (n)

2 4 6 8

Traces

FP (ms)

Trials

1500

500

1000

500

500 IA

500 1000 1500

Foreperiod (ms)

Foreperiod (ms)

Preparation (a.u.)

Short-Term Dynamics

Long-Term Dynamics

500

1000 1500

500

1000 1500

Exp

Uniform

Anti-Exp

FP

_n-1

(ms)

500

1000

1500

fMTP vs Hazard

fMTP and the Hazard function make different

predictions regarding the Gaussian FP distribution

Distribution

Across Blocks

fMTP

Hazard

Function

Probabili

ty

0.80

0.60

0.40

0.20

Foreperiod (ms)

1300 1950 2600Catch

Foreperiod (ms)

1300

1950

2600

400

375

350

325

Experimental

Assessment

Reacti

on Time (ms)

Empirical data

Hazard

fMTP

Reacti

on Time (ms)

The Story

_{FP Effects}

fMTP's Dynamics

Foreperiod (ms)

500 1000 1500

Anti-exp

Uniform

Exp

400

375

350

325

Reacti

on Time (ms)

FP

_n-1

(ms)

400

800

1200

1600

1000

500

1500

Foreperiod (ms)