STRIDE: Learnable Stepwise Language Feedback for LLM Reasoning
Open paper
Benchmark
Model
Rank
Results
math-word-problem-solving-on-math
STRIDE-Llama-8B
–
Accuracy: 70.4
math-word-problem-solving-on-math
STRIDE-Qwen-7B
–
Accuracy: 84.6
Rank counts only results with a code link.