Introspective Learning : A Two-Stage Approach for Inference in Neural Networks

2022-09-17 · Mohit Prabhushankar, Ghassan AlRegib

In this paper, we advocate for two stages in a neural network's decision making process. The first is the existing feed-forward inference framework where patterns in given data are sensed and associated with previously learned patterns. The second stage is a slower reflection stage where we ask the network to reflect on its feed-forward decision by considering and evaluating all available choices. Together, we term the two stages as introspective learning. We use gradients of trained neural networks as a measurement of this reflection. A simple three-layered Multi Layer Perceptron is used as the second stage that predicts based on all extracted gradient features. We perceptually visualize the post-hoc explanations from both stages to provide a visual grounding to introspection. For the application of recognition, we show that an introspective network is 4% more robust and 42% less prone to calibration errors when generalizing to noisy data. We also illustrate the value of introspective networks in downstream tasks that require generalizability and calibration including active learning, out-of-distribution detection, and uncertainty estimation. Finally, we ground the proposed machine introspection to human introspection for the application of image quality assessment.

📄 PDF Abstract BibTeX arXiv:2209.08425

Code (1)

olivesgatech/introspective-learning 공식 구현 pytorch

Tasks

Active Learning Decision Making Image Quality Assessment Out-of-Distribution Detection Visual Grounding Vocal Bursts Valence Prediction

IntroVAE: Introspective Variational Autoencoders for Photographic Image Synthesis

2018-07-17 · NeurIPS 2018 12 · Huaibo Huang, Zhihang Li, Ran He, Zhenan Sun 외

We present a novel introspective variational autoencoder (IntroVAE) model for synthesizing high-resolution photographic images. IntroVAE is capable of self-evaluating the quality of its generated samples and improving it…

Image Generation

Neuro-symbolic Natural Logic with Introspective Revision for Natural Language Inference

2022-03-09 · Yufei Feng, Xiaoyu Yang, Xiaodan Zhu, Michael Greenspan

We introduce a neuro-symbolic natural logic framework based on reinforcement learning with introspective revision. The model samples and rewards specific reasoning paths through policy gradient, in which the introspectiv…

Natural Language Inference reinforcement-learning Reinforcement Learning (RL)Systematic Generalization

Introspective Diffusion Language Models

2026-04-13 · Yifan Yu, Yuqing Jian, Junxiong Wang, Zhongzhu Zhou 외 arxiv

Diffusion language models promise parallel generation, yet still lag behind autoregressive (AR) models in quality. We stem this gap to a failure of introspective consistency: AR models agree with their own generations, w…

Introspective X Training: Feedback Conditioning Improves Scaling Across all LLM Training Stages

2026-05-19 · Brandon Cui, Ximing Lu, Jaehun Jung, Syeda Nahida Akter 외 arxiv

We tackle the question of how to scale more efficiently across the many, ever-growing stages of current LLM training pipelines. Our guiding intuition stems from the fact that the dynamics of later stages of the pipeline,…

Reinforcement Learning

Mechanisms of Introspective Awareness

2026-03-22 · Uzay Macar, Li Yang, Atticus Wang, Peter Wallich 외 arxiv

Recent work has shown that LLMs can sometimes detect when steering vectors are injected into their residual stream and identify the injected concept -- a phenomenon termed "introspective awareness." We investigate the me…