Work

HFT at BEA 2026 Shared Task 2: Blunt-Edge Models for Hybrid Grading

Year: 2026

Type: conference-paper

Abstract: Open-source LLMs with simple, zero-shot prompts are at best middling graders on the BEA 2026 Automated Grading Shared Taskblunt-edge models, in fact.However, they are good enough to support human graders and save them time.We demonstrate the application of a hybrid grading approach that first transparently defines the success criteria and then pairs a zero-shot LLM grader with human review.The hybrid approach outperforms the LLM grader on its own and has the added advantage of keeping the human in the loop.

Source: Workshop on Innovative Use of NLP for Building Educational Applications

DOI: 10.18653/v1/2026.bea-1.86

Author: Ulrike Pado

Institution: Stuttgart Technical University of Applied Sciences

Language: English

Topic: Natural Language Processing Techniques

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

SDG: Quality education

Open Access status: hybrid

OpenAlex ID: W7166800958