AI Paper Atlas

Research map / Large language models / Language, safety and interpretability

A Compress-Expand Framework for Automatic Lesson Plan Generation

Shuyan Huang, Ying Zheng, Xiaoli Zeng, Zitao Liu

AAAI 2026 · Large language models › Language, safety and interpretability · approach: LLM unlearning and benchmarks · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add textbook expander (ablation)CE-LessonPlan · Lesson plan generationROUGE-1: 11.606 -> 14.849Improved
add lesson plan compressor (ablation)CE-LessonPlan · Lesson plan generationROUGE-1: 18.143 -> 14.849Hurt
add lesson plan compressor (ablation)CE-LessonPlan · Lesson plan generationHuman evaluation average score: 2.273 -> 2.276No clear effect
add compressor and expander to the SFT+RAG generator (ablation)CE-LessonPlan · Lesson plan generationROUGE-1: 7.334 -> 14.849Improved
add three retrieved lesson plans as few-shot examplesGPT-4o · Lesson plan generationROUGE-1: 6.708 -> 8.779Improved
add textbook information to the promptGPT-4o · Lesson plan generationROUGE-1: 6.708 -> 14.016Improved
add 3-shot retrieved lesson plans to textbook-enhanced promptGPT-4o · Lesson plan generationROUGE-1 and ROUGE-L: 14.016 -> 14.324 ROUGE-1; 12.775 -> 12.662 ROUGE-LMixed
raise ensemble coefficient alpha from 0.5 up to 0.8, then to 0.9 (ablation)CE-LessonPlan · Lesson plan generationROUGE-1 and ROUGE-LMixed
use SFT lesson plan model instead of Qwen2.5-72B as generator (ablation)CE-LessonPlan · Lesson plan generationROUGE-1: 14.511 -> 14.849Improved
use SFT lesson plan model instead of Qwen2.5-72B as compressor and expander (ablation)CE-LessonPlan · Lesson plan generationROUGE-1: 11.964 -> 14.849Improved
use compress-expand framework instead of Self-Critiquelesson plan generator · Lesson plan generationROUGE-1: 13.317 -> 14.849Improved
use compress-expand framework instead of Self-Critiquelesson plan generator · Lesson plan generationHuman MAT score: 2.605 -> 1.933Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy