List your company
ResearchFormerly WizWand

Mathematical Reasoning on iGSM

54.25AccuracySupervised
11.0922.29533.544.705May 7, 2026
Updated 5mo ago

Evaluation results

MethodLinks
SupervisedLLM=qwen2.5-7b-instruct2026.0554.25
LOVER (sum)LLM=qwen2.5-7b-instruct2026.0552
Majority VotingLLM=qwen2.5-7b-instruct2026.0551
CoT-Decoding (sum)LLM=qwen2.5-7b-instruct2026.0551
LOVER (max)LLM=qwen2.5-7b-instruct2026.0548.25
GreedyLLM=qwen2.5-7b-instruct2026.0546.5
SupervisedLLM=mistral-7b-instruct-v0.32026.0544.5
SupervisedLLM=llama-3.1-8b-instruct2026.0543.75
LOVER (sum)LLM=llama-3.1-8b-instruct2026.0542.5
LOVER (sum)LLM=mistral-7b-instruct-v0.32026.0541.25
CoT-Decoding (sum)LLM=llama-3.1-8b-instruct2026.0540.5
Majority VotingLLM=llama-3.1-8b-instruct2026.0539
LOVER (max)LLM=llama-3.1-8b-instruct2026.0539
CoT-Decoding (sum)LLM=mistral-7b-instruct-v0.32026.0536.75
GreedyLLM=llama-3.1-8b-instruct2026.0535
Majority VotingLLM=mistral-7b-instruct-v0.32026.0534.5
CoT-Decoding (max)LLM=qwen2.5-7b-instruct2026.0531.5
LOVER (max)LLM=mistral-7b-instruct-v0.32026.0528.75
CoT-Decoding (max)LLM=llama-3.1-8b-instruct2026.0523.25
GreedyLLM=mistral-7b-instruct-v0.32026.0519
CoT-Decoding (max)LLM=mistral-7b-instruct-v0.32026.0512.75