Saved in:
Bibliographic Details
Main Authors: Sadhu, Saisab, Dhor, Ashim
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2512.22496
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909976789778432
author Sadhu, Saisab
Dhor, Ashim
author_facet Sadhu, Saisab
Dhor, Ashim
contents Large Language Models (LLMs) are increasingly deployed as automated tutors to address educator shortages; however, they often fail at pedagogical reasoning, frequently validating incorrect student solutions (sycophancy) or providing overly direct answers that hinder learning. We introduce Hierarchical Pedagogical Oversight (HPO), a framework that adapts structured adversarial synthesis to educational assessment. Unlike cooperative multi-agent systems that often drift toward superficial consensus, HPO enforces a dialectical separation of concerns: specialist agents first distill dialogue context, which then grounds a moderated, five-act debate between opposing pedagogical critics. We evaluate this framework on the MRBench dataset of 1,214 middle-school mathematics dialogues. Our 8B-parameter model achieves a Macro F1 of 0.845, outperforming GPT-4o (0.812) by 3.3% while using 20 times fewer parameters. These results establish adversarial reasoning as a critical mechanism for deploying reliable, low-compute pedagogical oversight in resource-constrained environments.
format Preprint
id arxiv_https___arxiv_org_abs_2512_22496
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Hierarchical Pedagogical Oversight: A Multi-Agent Adversarial Framework for Reliable AI Tutoring
Sadhu, Saisab
Dhor, Ashim
Multiagent Systems
Artificial Intelligence
Large Language Models (LLMs) are increasingly deployed as automated tutors to address educator shortages; however, they often fail at pedagogical reasoning, frequently validating incorrect student solutions (sycophancy) or providing overly direct answers that hinder learning. We introduce Hierarchical Pedagogical Oversight (HPO), a framework that adapts structured adversarial synthesis to educational assessment. Unlike cooperative multi-agent systems that often drift toward superficial consensus, HPO enforces a dialectical separation of concerns: specialist agents first distill dialogue context, which then grounds a moderated, five-act debate between opposing pedagogical critics. We evaluate this framework on the MRBench dataset of 1,214 middle-school mathematics dialogues. Our 8B-parameter model achieves a Macro F1 of 0.845, outperforming GPT-4o (0.812) by 3.3% while using 20 times fewer parameters. These results establish adversarial reasoning as a critical mechanism for deploying reliable, low-compute pedagogical oversight in resource-constrained environments.
title Hierarchical Pedagogical Oversight: A Multi-Agent Adversarial Framework for Reliable AI Tutoring
topic Multiagent Systems
Artificial Intelligence
url https://arxiv.org/abs/2512.22496