~/tedtes/writing/math-foundations-for-reasoning-models

~/writing/math-foundations-for-reasoning-models $ ls

Reasoning Models as Representation Transformation

Intermediate Representation

An intermediate representation is a temporary transformed state between the input and final output.

~/writing/math-foundations-for-reasoning-models $ cat topic.md

Intermediate Representation

Definition

An intermediate representation is a representation produced between the input and the final output.

Simple form:

input representation -> intermediate representation -> output representation

It is not the final answer.

It is a temporary state used during computation.

Intermediate representations carry partial structure during reasoning.


Why Intermediate Representations Matter

Reasoning often cannot jump directly from input to output.

The model may need to hold:

partial results
constraints
possible steps
relationships
uncertainty

Intermediate representations let the model preserve and transform this information.

Multi-step tasks need useful states between start and finish.


Visible vs Hidden Intermediate Representations

Some intermediate representations are visible.

Example:

written scratchpad text
tool output
draft answer

Others are hidden inside the model.

Example:

hidden states across layers
internal activations
contextual embeddings

Both can affect the final output.

Intermediate representation can be external text or internal model state.


Not Necessarily Human-Readable

Internal representations do not need to look like human reasoning.

They may be distributed across many dimensions.

They may not map cleanly to a sentence or symbolic step.

useful for model computation
not directly readable by humans

A representation can be computationally useful without being interpretable.


Connection To State

Earlier, state was the current description of a system.

An intermediate representation is part of the model's changing state during computation.

state at step t -> state at step t + 1

Reasoning can be viewed as a sequence of representational states.

Intermediate representations are reasoning states in motion.


Final Summary

  • An intermediate representation exists between input and output.
  • It carries partial information during computation.
  • It may be visible or hidden.
  • It does not need to be human-readable.
  • Reasoning can be understood as changing intermediate representations.

~/writing/math-foundations-for-reasoning-models $ open

Continue the series.

Multi-Step Transformation