mlsanityData to 14 Oct 2026

Automated IEP Generation from Traditional Chinese Parent-Teacher Interviews via Corpus-Grounded Feature Diffusion

arXiv 2606.09603 ↗First posted 8 Jun 2026cs.CLcited by 0

Not on the map. Why?

Abstract (arXiv)

Writing Individualized Education Programs (IEPs) is a high-labor, knowledge-intensive document burden; English-language research has demonstrated that generative AI can significantly reduce drafting time, yet automated IEP generation in Traditional Chinese remains virtually unexplored due to domain data scarcity, strict privacy regulations, and the absence of local evaluation benchmarks. We propose a low-resource fine-tuning pipeline centered on Corpus-Grounded Feature Diffusion (CGFD): (1) 25 dual-expert high-score seed transcripts are selected via a tau threshold with flag-aware score caps; (2) a FeatureProfile (sentence length, structure, quantification templates) is extracted from seeds and injected into LLM prompts alongside Verbalized-Sampling-style diversity control to drive diffusion; (3) 15 expert gold seeds are used as diffusion anchors, targeting 585 samples; 567 valid diffusion samples are obtained, yielding a 582-sample training set used to fine-tune Breeze-7B with QLoRA; (4) schema-constrained inference via Grammar-Constrained Decoding (GCD) enforces a hierarchical SMART Goal Ladder schema at inference time. Ablation results on a 55-sample schema stress set reveal an unexpected finding: GCD is counterproductive under Traditional Chinese token budgets -- the no-GCD path achieves 100% schema pass rate at 34% lower median latency, outperforming GCD on both reliability and speed. On the n=10 formal hold-out, the no-GCD inference path achieves BERTScore F1 = 0.779, exceeding GPT-5.4 (0.726), DeepSeek-V3.2 (0.703), Gemini-3-Flash-Preview (0.703), and Llama-4-Maverick (0.700) zero-shot baselines while maintaining fully local, air-gapped inference. This system addresses a gap in Traditional Chinese special-education NLP and offers a scalable, privacy-preserving local inference solution under an industrial engineering paradigm.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 18 of its 31 references that are in the map. First posted on arXiv 8 Jun 2026.

Builds on

18 of 31 references.

References · 18

  1. 2019BERTScore: Evaluating Text Generation with BERTcited by 5,512
  2. 2022Self-Instruct: Aligning Language Models with Self-Generated Instructionscited by 2,724
  3. 2023WizardLM: Empowering large pre-trained language models to follow complex instructionscited by 1,125
  4. 2024Self-Preference Bias in LLM-as-a-Judgecited by 258
  5. 2023Grammar-Constrained Decoding for Structured NLP Tasks without Finetuningcited by 197
  6. 2025Verbalized Sampling: How to Mitigate Mode Collapse and Unlock LLM Diversitycited by 105
  7. 2025SPICE: Self-Play In Corpus Environments Improves Reasoningcited by 83
  8. 2025Flexible and Efficient Grammar-Constrained Decodingcited by 43
  9. 2025Beyond the Surface: Measuring Self-Preference in LLM Judgmentscited by 31
  10. 2024Automatic Instruction Evolving for Large Language Modelscited by 26
  11. 2025What and Whose Knowledge? Measuring Epistemic Diversity in Large Language Modelscited by 14
  12. 2026Counting on Consensus: Selecting the Right Inter-annotator Agreement Metric for NLP Annotation and Evaluationcited by 13
6 more
  1. 2026Quantifying and Mitigating Self-Preference Bias of LLM Judgescited by 10
  2. 2025Asking a Language Model for Diverse Responsescited by 9
  3. 2024Breeze-7B Technical Reportcited by 4
  4. 2025The Breeze 2 Herd of Models: Traditional Chinese LLMs Based on Llama with Vision-Aware and Function-Calling Capabilitiescited by 4
  5. 2025SQaLe: A Large Text-to-SQL Corpus Grounded in Real Schemascited by 3
  6. 2026Limits of n-gram Style Control for LLMs via Logit-Space Injectioncited by 1

Built on by

Not cited yet.