From Execution to Education: A Bloom-Aligned Framework for Measuring Educational Control in LLMs

Jul 16, 2026

Introduces a Bloom-aligned framework to measure educational control in Large Language Models (LLMs), defined as preserving a task’s instructional intent while shifting cognitive demand toward targeted learning objectives. The framework is applied to computer science programming tasks to quantify the gap between solving and adapting for learners. Using a revised Bloom’s Taxonomy as an operational scale, two interventions are tested: general difficulty control and Bloom-level targeting. Evaluations on 2,520 tasks across three benchmarks show strong ability to raise cognitive demand but persistent difficulty lowering it.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages24

Reading Time1 h 0 min

Abstract

Introduction

Motivation and problem framing

Limitations of execution-only benchmarks

Prior work and the research gap

Contributions and evaluation overview

arXiv :2607 .08009v 1 [ cs .CL] 9 Jul 2026 From Execution to Education: A Bloom-Aligned Framework for Measuring Educational Control in LLMs Yi Zhang & Julia Rayz School of Applied Creative Computing Purdue University West Lafayette, IN 47907, USA {zhan3050,[jtaylor1](jtaylor1}@purdue.edu)[}](jtaylor1}@purdue.edu)[@purdue.edu](jtaylor1}@purdue.edu) Abstract We introduce a Bloom-aligned framework for measuring educational control in Large Language Models (LLMs): the ability to preserve a task’s instructional intent while shifting its cognitive demand toward specified learning objectives. We apply this framework to programming tasks in computer science education to study the gap between solving tasks and adapting them for learners. Using revised Bloom’s Taxonomy as an operational scale of cognitive demand, we evaluate two intervention settings: general difficulty control, where models are asked to make tasks harder or easier, and Bloom’s control, where models are asked to target higher or lower Bloom’s levels. We evaluate a matched Qwen3-Next model pair, comparing Qwen3-Next-80B-A3B-Instruct with Qwen3-Coder-Next across 2,520 tasks from three benchmarks. The framework reveals a robust directional asymmetry: both models reliably increase cognitive demand, but struggle to lower it. We further characterize these outcomes with semantic-delta clustering and layer-wise Fisher’s Discriminant Ratio probing. Within this controlled comparison, the general model shows clearer middle-layer separability for both general difficulty and Bloom-control contrasts, whereas the coder model shows weaker separability for general difficulty and a deeper peak for Bloom-control contrasts. These results show that strong execution performance does not automatically entail Bloom-aligned educational control. 1 Introduction Recent advancements in Large Language Models (LLMs) have demonstrated significant success in extending their capabilities to complex software engineering tasks. As these models become deeply integrated into real-world applications, they are increasingly deployed beyond code completion to act as collaborative educational tools (Hou et al., 2024) and intelligent tutors (HKUDS, 2025; Yu et al., 2026) . However, excelling at functional execution benchmarks illustrates that a model can be a proficient problem solver without being an effective educator. The Zone of Proximal Development (ZPD) (Vygotsky, 1980) provides useful background for this distinction: education requires dynamically adjusting difficulty so that tasks remain challenging enough to encourage learning without becoming unproductive. We therefore study educational control: the task-level ability to preserve instructional intent while shifting cognitive demand toward specified learning objectives. In this paper, we operationalize educational control with revised Bloom’s Taxonomy (Anderson & Krathwohl, 2001), which provides an ordered structure of cognitive processes. Programming tasks provide a controlled testbed for observing the gap between execution performance and educational adaptation. To measure LLM proficiency in software engineering, the research community has developed a vast and diverse ecosystem of benchmarks, ranging from static, function-level evaluations like HumanEval (Chen et al., 2021) to complex, repository-level frameworks such as SWE-bench (Jimenez et al., 2024) . While these execution metrics are essential for ensuring that generated code is valid, they are pedagogically shallow because they treat code generation as a pass-or-fail outcome and ignore the underlying cognitive load required to solve the problem. When LLMs are applied directly to computer science education, optimizing solely for immediate, functionally correct solutions frequently causes an “illusion of learning”(Prather et al., 2024) . To leverage these models as educational tools, we must look beyond standard execution benchmarks and implement extra considerations to ensure

FAQs

What does “educational control” mean in this paper?

Educational control is the task-level ability to preserve a task’s instructional intent while shifting its cognitive demand toward specified learning objectives.

How is cognitive demand operationalized for measurement?

The paper uses a revised Bloom’s Taxonomy as an ordered scale of cognitive processes, treating Bloom levels as measurable targets for cognitive demand shifts.

What interventions are evaluated to test educational control?

Two settings are studied: general difficulty control (asking for harder/easier tasks) and Bloom’s control (targeting higher/lower Bloom levels).

What key behavioral finding emerges from the experiments?

Both evaluated models reliably increase cognitive demand, but they often struggle to decrease it, indicating a directional asymmetry in Bloom-aligned educational control.

Jul 16, 2026

Upload by Genevieve