Thought Flow Nets: From Single Predictions to Trains of Model Thought

TL;DR

Thought Flow Nets introduce a sequence of iterative predictions inspired by human reasoning, enabling models to self-correct and improve accuracy in question answering tasks while enhancing user perception of naturalness and intelligence.

Contribution

The paper proposes a novel thought flow mechanism with self-correction for models, allowing multiple prediction steps and improving performance and interpretability.

Findings

Model can correct its own predictions through thought flow.

Thought flow improves question answering accuracy.

Users perceive thought flows as more natural and intelligent.

Abstract

When humans solve complex problems, they typically create a sequence of ideas (involving an intuitive decision, reflection, error correction, etc.) in order to reach a conclusive decision. Contrary to this, today's models are mostly trained to map an input to one single and fixed output. In this paper, we investigate how we can give models the opportunity of a second, third and $k$-th thought. Taking inspiration from Hegel's dialectics, we propose the concept of a thought flow which creates a sequence of predictions. We present a self-correction mechanism that is trained to estimate the model's correctness and performs iterative prediction updates based on the correctness prediction's gradient. We introduce our method at the example of question answering and conduct extensive experiments that demonstrate (i) our method's ability to correct its own predictions and (ii) its potential to notably improve model performances. In addition, we conduct a qualitative analysis of thought flow correction patterns and explore how thought flow predictions affect human users within a crowdsourcing study. We find that (iii) thought flows enable improved user performance and are perceived as more natural, correct, and intelligent as single and/or top-3 predictions.