List your company

Fact Generation on Wiki ExAnte (test)

11.76Leakage RateTCFT
10.993216.169121.34526.5209May 14, 2026

Evaluation results

MethodLinks
TCFTBackbone=Qwen-2.5-14B-Instruct, Evaluation Protocol=Ours2026.0511.7673.81
GPT-5.4Model=GPT-5.42026.0513.2881.37
TCFTBackbone=Qwen-2.5-7B-Instruct, Evaluation Protocol=Ours2026.0514.0466.24
Gemini-3.1-Pro-PreviewModel=Gemini-3.1-Pro-Preview2026.0515.7579.82
SFTBackbone=Qwen-2.5-14B-Instruct, Evaluation Protocol=SFT2026.0516.3271.01
Claude-Opus-4.6Model=Claude-Opus-4.62026.0516.8681.3
Self-verificationBackbone=Qwen-2.5-14B-Instruct, Evaluation Protocol=Self-verification2026.0520.4970.73
SFTBackbone=Qwen-2.5-7B-Instruct, Evaluation Protocol=SFT2026.0520.7962.18
Zero-shotBackbone=Qwen-2.5-7B-Instruct, Evaluation Protocol=Zero-shot2026.0521.8261.67
Self-verificationBackbone=Qwen-2.5-7B-Instruct, Evaluation Protocol=Self-verification2026.0522.0162.81
Chain-of-thoughtBackbone=Qwen-2.5-7B-Instruct, Evaluation Protocol=Chain-of-thought2026.0522.7762.35
Chain-of-thoughtBackbone=Qwen-2.5-14B-Instruct, Evaluation Protocol=Chain-of-thought2026.0523.1567.86
Zero-shotBackbone=Qwen-2.5-14B-Instruct, Evaluation Protocol=Zero-shot2026.0523.3469.49
Few-shotBackbone=Qwen-2.5-7B-Instruct, Evaluation Protocol=Few-shot2026.0526.7659.96
Few-shotBackbone=Qwen-2.5-14B-Instruct, Evaluation Protocol=Few-shot2026.0530.9366.68