Claim-level rating prediction on Insq benchmark leaderboard · SOTA2 Research

Claim-level rating prediction on Insq

0.726CIG ScoreQwen3-4B

Evaluation Results

Method	Links
Qwen3-4B2026.04		0.726	0.617	0.509	0.569	0.605
Gemini-2.5-Flash2026.04		0.689	0.652	0.71	0.453	0.626
Gemini-2.5-Flash2026.04		0.677	0.621	0.648	0.465	0.603
Qwen3-4B2026.04		0.658	0.618	0.528	0.593	0.599
Gemini-2.5-Flash2026.04		0.611	0.628	0.579	0.458	0.569
Qwen3-4B2026.04		0.601	0.673	0.494	0.586	0.589
Qwen3-4B2026.04		0.597	0.656	0.589	0.597	0.61
GPT-5-Mini2026.04		0.577	0.53	0.551	0.473	0.533
GPT-5-Mini2026.04		0.557	0.599	0.52	0.472	0.537
Gemini-2.5-Flash2026.04		0.554	0.6	0.502	0.467	0.531
GPT-5-Mini2026.04		0.546	0.806	0.547	0.489	0.597
Gemini-2.5-Flash2026.04		0.543	0.653	0.615	0.496	0.577
Qwen3-4B2026.04		0.536	0.812	0.569	0.626	0.636
GPT-5-Mini2026.04		0.501	0.697	0.535	0.467	0.55
GPT-5-Mini2026.04		0.487	0.726	0.529	0.484	0.557
Gemini-2.5-Pro2026.04		0.484	0.506	0.556	0.523	0.517
Gemini-2.5-Pro2026.04		0.48	0.519	0.551	0.501	0.513
Gemini-2.5-Pro2026.04		0.479	0.476	0.566	0.513	0.509
Gemini-2.5-Pro2026.04		0.471	0.601	0.61	0.518	0.55
GPT-52026.04		0.47	0.7	0.436	0.539	0.536
GPT-52026.04		0.459	0.567	0.427	0.526	0.495
GPT-52026.04		0.459	0.637	0.44	0.536	0.518
GPT-52026.04		0.457	0.587	0.452	0.529	0.506
Gemini-2.5-Pro2026.04		0.431	0.58	0.578	0.512	0.525
GPT-52026.04		0.411	0.577	0.444	0.541	0.493