> ## Documentation Index
> Fetch the complete documentation index at: https://docs.thrindex.com/llms.txt
> Use this file to discover all available pages before exploring further.

# E0106 — DimensionMismatch (compiler)

> Two consecutive layers have incompatible dimensions.

<Warning>E0106: layer dimension mismatch</Warning>

## What happened

During the compiler's validate pass, it found that layer `N` expects `expected` input features, but layer `N-1` produces `got` output features. The two layers cannot be connected.

## Why

The most common cause: you changed the size of one layer but forgot to update the adjacent layer.

For example:

```python theme={null}
# Bug: Dense(512, 20) can't connect to LIF that was set up for 256 outputs
model = snn.Sequential(
    snn.Dense(700, 256),   # outputs 256
    snn.LIF(tau_mem=20.0, threshold=0.3),
    snn.Dense(512, 20),    # expects 512 but gets 256
    snn.LIF(tau_mem=20.0, threshold=0.3),
)
```

## How to fix

1. Check each `Dense` layer — `out_features` of layer `N` must match `in_features` of layer `N+2` (accounting for the LIF layer in between).
2. Use a consistent naming pattern, e.g. `hidden_size = 512`, and reference the variable throughout:

```python theme={null}
hidden_size = 512
model = snn.Sequential(
    snn.Dense(700, hidden_size),
    snn.LIF(tau_mem=20.0, threshold=0.3),
    snn.Dense(hidden_size, 20),
    snn.LIF(tau_mem=20.0, threshold=0.3),
)
```

## Example

```
E0106: layer dimension mismatch at layer 2
Why: layer 2 declares 512 input features but layer 1 produces 256 output features.
Fix: recheck the architecture — input/output sizes must chain correctly.
Docs: https://docs.thrindex.com/errors/E0106
```