PulseAugur
EN
LIVE 09:56:50

New corpus and framework evaluate LLM writing feedback quality

By PulseAugur Editorial · [2 sources] ·

Researchers have introduced SEFORA, a new corpus designed to capture how instructors provide feedback on student writing, alongside UniMatch, an evaluation framework for assessing the quality of AI-generated feedback. SEFORA contains over 8,000 instructor annotations across 564 student essay drafts. The UniMatch framework measures the semantic correspondence and alignment of AI feedback units against instructor-derived criteria. Experiments using UniMatch showed that current LLMs struggle to produce feedback that aligns with instructor priorities, with performance degrading as more feedback is generated, and no tested configuration exceeding a 0.4 F1 score. AI

IMPACT This research highlights current limitations in LLM writing feedback, suggesting a need for improved alignment with human instructor priorities.

RANK_REASON The cluster describes a new academic paper introducing a corpus and evaluation framework for LLM-generated writing feedback. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CL →

AI-generated summary · Google Gemini · from 2 sources. How we write summaries →

New corpus and framework evaluate LLM writing feedback quality

COVERAGE [2]

  1. arXiv cs.AI TIER_1 English(EN) · Shayan Peyghambari Oskoui, Norah Almousa, Zhaoyi Joey Hou, Carolina Gustafson, Gayle Rogers, Raquel Coelho, Diane Litman, Xiang Lorraine Li ·

    SEFORA: Student Essays with Feedback Corpus and LLM Feedback Evaluation Framework

    arXiv:2607.00274v1 Announce Type: cross Abstract: Effective writing feedback is among the strongest drivers of student learning, yet producing it at scale is labor-intensive. LLMs offer a natural path to scaling writing support, but two gaps stand in the way: few public corpora c…

  2. arXiv cs.CL TIER_1 English(EN) · Xiang Lorraine Li ·

    SEFORA: Student Essays with Feedback Corpus and LLM Feedback Evaluation Framework

    Effective writing feedback is among the strongest drivers of student learning, yet producing it at scale is labor-intensive. LLMs offer a natural path to scaling writing support, but two gaps stand in the way: few public corpora capture how instructors actually deliver feedback i…

RELATED ENTITIES

RELATED TOPICS