Protocol

Mouse visual contrast wheel task with reward-value blocks

Head-fixed mice report the side of a lateralized visual contrast stimulus with a response wheel while reward value differs between sides and switches across blocks. This variant separates reward-value learning from the IBL stimulus-prior manipulation, even though both use wheel-based visual 2AFC behavior.

Task structure

Stimulus, choice, feedback, and training as declared in the protocol record.

Stimulus

evidence_schedule
A sinusoidal grating or similar visual stimulus appears on the left or right monitor at variable contrast.
evidence_type
static
modalities
visual
notes
—
units
percent contrast
variables
contrast, stimulus side

Choice

action_mapping
Wheel turns report the side of the visual stimulus.
alternatives
left, right
choice_type
2afc
notes
—
response_modalities
wheel

Feedback

feedback_type
mixed
notes
Reward-value block state is an operational task variable and should remain separate from sensory confidence or inferred decision confidence.
penalty
Error feedback or timeout for incorrect choices, source-dependent.
reward
Water reward for correct choices; reward magnitude or value can be larger on one side and switch across blocks.

Training

notes
—
stages
wheel shaping, visual discrimination, reward-value block training

Trial timing

  1. quiescence
    protocol-specific

    Mouse holds the wheel still before stimulus onset.

  2. stimulus_and_response
    response-limited

    Visual stimulus is presented while the mouse reports with the wheel.

  3. feedback
    outcome-dependent

Expected analyses

  • psychometric curve by contrast and reward-value block
  • choice updating after rewarded and unrewarded trials
  • reinforcement-learning model with sensory-confidence terms
  • neural correlates of pending reward and prediction error

Interpretive claims

  • reward-value learning under sensory uncertainty
    Confidence: medium · ref.lak-2020-dopaminergic-prefrontal

    Sensory contrast, reward value, and inferred confidence are separable quantities and should not be collapsed into a single evidence axis.

Open questions

  • Identify whether public trial-level behavioral tables exist for the Lak visual contrast reward-value studies.
  • Decide whether the cross-species Lak eLife analyses require separate human and mouse source-specific protocol records.