Scroll's architecture separates sequencing, proving, and settlement. Execution nodes process both user transactions and messages originating from the Ethereum bridge. Rollup nodes group the resulting activity and publish data; provers construct evidence that the execution followed the relevant rules. Ethereum hosts the bridge and rollup contracts that verify proofs and record settlement. This division allows expensive application computation to happen away from Ethereum while retaining an Ethereum-based verification path.

It does not mean the sequencer, prover, and settlement contract are the same component.

The documented rollup process proceeds through execution, batching and data commitment, and proof generation and finalization. A commitment transaction publishes batch data; a later finalization transaction verifies the corresponding proof. An explorer can therefore show an L2 transaction before every settlement step is complete. For example, a support investigation into a delayed transfer should determine whether execution failed, the batch has not been committed, or proof finalization remains pending.

These are different failure or waiting states, with different evidence in the chain's records.

Scroll's withdrawal gateways coordinate the Layer 2 action with a message that releases or represents value on Ethereum after the required verification. The relevant gateway depends on the token and bridge implementation. A token balance on Scroll is therefore not an Ethereum-mainnet balance merely because both are displayed under the same wallet address. Checking the source chain, destination chain, asset contract, and message status prevents that confusion.

A validity proof verifies the defined transition; correctness of the token gateway and availability of the supporting infrastructure remain necessary parts of the transfer.