List your company

Financial Question Answering on FinQA

83.46AccuracyTaNOS
-0.977620.943742.86564.7863Mar 20, 2025May 25, 2025Jul 31, 2025Oct 5, 2025Dec 11, 2025Feb 15, 2026Apr 23, 2026

Evaluation results

MethodLinks
TaNOSTrain Data=FinQA2026.0483.46
Formula-R1-PlusBackbone=Qwen 2.5-Coder 7B, Evaluation Protocol=Finetuning-Based2025.0580.47
Fin-R1Parameters=7B2025.0376
SMCS2025.0775.24
Self-MoA2025.0774.11
QwQ-32B2025.0773.67
GPT-4.12025.0772.97
Qwen2.5-32B-InstructParameters=32B2025.0372
DeepSeek-R1Parameters=671B2025.0371
DeepSeek-R1-Distill-Qwen-32BParameters=32B2025.0370
RL-F1Reward=RL-F1, Rollout-Group Rejection (Rubric)=false, Rollout-Group Rejection (Variance)=false2025.0669
SFTTrain Data=FinQA2026.0468.72
Full DROReward=DRO, Rollout-Group Rejection (Rubric)=true, Rollout-Group Rejection (Variance)=true2025.0668.4
Qwen2.5-14B-InstructParameters=14B2025.0368
DeepSeek-R1-Distill-Llama-70BParameters=70B2025.0368
R3Reward=R3, Rollout-Group Rejection (Rubric)=false, Rollout-Group Rejection (Variance)=false2025.0667.1
Avg Prob (RLPR)Reward=Avg Prob (RLPR), Rollout-Group Rejection (Rubric)=false, Rollout-Group Rejection (Variance)=false2025.0666.8
TableMasterBackbone=GPT-4o-mini, Evaluation Protocol=Prompting-Based2025.0566.4
Avg Logprob (VeriFree)Reward=Avg Logprob (VeriFree), Rollout-Group Rejection (Rubric)=false, Rollout-Group Rejection (Variance)=false2025.0666.4
Formula-R1Backbone=Qwen 2.5-Coder 7B, Evaluation Protocol=Finetuning-Based2025.0562.16
DeepSeek-R1-Distill-Qwen-14BParameters=14B2025.0362
Qwen2.5-7B-InstructParameters=7B2025.0360
BaseReward=Base2025.0658.1
DeepSeek-R1-Distill-Qwen-7BParameters=7B2025.0355
SFTTrain Data=MultiHiertt2026.0450.51
TabAFBackbone=Qwen 2.5-Coder 7B, Evaluation Protocol=Finetuning-Based, Distribution Setting=Out-of-distribution2025.0545.07
TableGPT2Backbone=Qwen 2.5 7B, Evaluation Protocol=Finetuning-Based, Distribution Setting=Out-of-distribution2025.0540.28
SFTTrain Data=NumReason-5002026.0427.44
TableLLMBackbone=Qwen 7B, Evaluation Protocol=Finetuning-Based, Distribution Setting=Out-of-distribution2025.058.63
TableLlamaBackbone=Llama-2 7B, Evaluation Protocol=Finetuning-Based, Distribution Setting=Out-of-distribution2025.052.27