Kaggle2025Hackathons
MAP: Charting Student Math Misunderstandings
Reading a student's written maths explanation and deciding whether it's correct, a slip, a named misconception, or none of those. Two stages: correctness first, then the misconception.
Sentence embeddingsPhi instructTwo-stage pipeline
Status: Draft.
Objective
Classify a student's written maths explanation as correct, a slip, a named misconception, or neither.
I've asked myself whether I'm actually bad at maths, so this one is a bit personal. Getting an answer wrong and holding a misconception are not the same thing, and I think that difference is the most interesting part of the task.
Approach
A two-stage pipeline, so the harder question only gets asked when it's needed:
- Stage 1: predict whether the answer is correct.
- Stage 2: for wrong answers only, detect the misconception using sentence embeddings plus a Phi instruct LLM.
Where it's at
- 🔄 Stage 1 correctness model
- 🔄 Stage 2 misconception detection
Posts
No posts for this project yet. Come back later.