← home
06 — competition

competition.

pushing the boundary of LLM mathematical reasoning.

Deep Learning Competition · LLM reasoning

2nd Solo8th Team

KAGGLE / SPRING 2026

Pushing the Boundary
of LLM Mathematical
Reasoning.

CSE 151B: Deep Learning, Spring 2026 Competition

Improving the mathematical reasoning capabilities of Qwen3-4B-Thinking, a four-billion-parameter language model. Competed solo on problems spanning high-school to graduate-level mathematics.