Read the equation
The update is y_new=0.99y_old+x−x_old. It retains most of the previous y and adds a displacement measured against the history value called x_old. When that value is the oldest element of the forty-four-sample buffer, this is a delayed-displacement accumulator rather than an instantaneous finite-difference velocity.
Terms & definitions
| Term | Meaning in this source |
|---|---|
| y | Accumulated velocity-like state |
| x | Current activation |
| x_old | Referenced history sample |
How it works
Read the definition from its right-hand expression to the quantity on the left. Its surrounding source records the intended input, state, or parameter context. Where the source leaves an operation unspecified, this page preserves that gap.
Assumptions & limits
Source recovery establishes that the expression appears in this record. Its mathematical validity, implementation behavior, and interpretation must be assessed separately. See the linked source and research notes for the checks available for this version.
Original context
Read the surrounding source passage
## 8.3 Velocity Update ``` y ← y + (x - x_old) - 0.01 · y ``` **Where**: - `x_old` = h[0] (oldest hidden state value) **Explanation**: Velocity tracks the rate of change with slight damping (0.01). This creates momentum in the agent's dynamics. ## 8.4 Drive Computation ``` drive = 3.0 · F_in · (1 - x) + 1.2 · y + 0.6 · c ``` **Where**: - `x` = current activation - `y` = velocity - `c` = coupling parameter **Explanation**: The drive combines: - Input-dependent excitation (stronger when x is low)
Exact original notation
y ← y + (x - x_old) - 0.01 · y