PulseExploreJournal ClubDebatesTrendingResearchersJournals
Instagram
HomeExploreJournal ClubTrending
Synapse
⌘+K
Synapse
May 4, 2026EPJ Web of Conferences0 citationsOpen Access

SymbAlign: A Hybrid Symbolic–Neural Alignment Framework for Automated Mathematical Solution Scoring

RJR. JohnsiBGBharadwaja Kumar Guntur

Key Points

  • This research aims to improve automated scoring of mathematical solutions by integrating symbolic and neural methods.
  • Developed a hybrid framework, SymbAlign++, utilizing both symbolic computation and neural semantic similarity.
  • Implemented an adaptive weighting mechanism to optimize the balance between symbolic and neural scores.
  • Evaluated performance on Hendrycks MATH dataset using metrics like R², QWK, and MSE.
  • SymbAlign++ significantly outperforms both symbolic-only and neural-only approaches.
  • Achieved optimal scoring across multiple mathematical solution categories with improved interpretability.
  • Performance metrics indicate a strong correlation between predicted and actual scores, enhancing scoring accuracy.

Abstract

Automated scoring of mathematical solutions is challenging due to the diversity of solution strategies and the need to assess both correctness and reasoning. We present SymbAlign++, a hybrid framework that combines symbolic computation and neural semantic similarity for stepwise evaluation of solutions. Symbolic similarity is computed using algebraic equivalence metrics, while neural similarity is captured through pre-trained language models. A per-category adaptive weighting mechanism (α) learns the optimal balance between symbolic and neural signals. Experiments were conducted on multiple categories from the Hendrycks MATH dataset, and the proposed SymbAlign++ achieves superior performance compared to symbolic-only and neural-only baselines. Performance was evaluated using various metrics, including R², Quadratic weighted kappa (QWK), MSE and correlation measures. The framework provides a robust, interpretable, and flexible approach for automated mathematical solution scoring supporting both procedural and semantic assessment.

Ask AI
Helpful
Bookmark
Share
View Full Paper

Cite This Study

Johnsi et al. (2026) studied this question.

synapsesocial.com/papers/69f836aa3ed186a739980dfahttps://doi.org/10.1051/epjconf/202636704009
Ask AI
Helpful
Bookmark
Share
View Full Paper