Uncertainty Attribution on UA-Bench Knowledge-intensive Tasks benchmark leaderboard · SOTA2 Research

Uncertainty Attribution on UA-Bench Knowledge-intensive Tasks

32.3AccuracyGemini 3 Flash

Evaluation Results

Method	Links
Gemini 3 Flash2026.04		32.3	72	29	50.5
LLaMA-4-Maverick2026.04		20.3	71	38.6	54.8
Qwen3-235B-A22B-Instruct-25072026.04		18	73.2	53.2	63.2
GPT-OSS 120B2026.04		17.3	72	48.5	60.2
GPT-4o mini2026.04		15.6	66.9	30.2	48.6
Qwen3-235B-A22B-Thinking-25072026.04		14.4	67.9	40.5	54.2
GPT-4o2026.04		10.4	78.2	66.6	72.4
GPT-OSS 20B2026.04		9.8	67.7	24.6	46.1
Qwen3-32B2026.04		8.8	73.1	65	69
Claude Sonnet 42026.04		8.3	74.2	67.4	70.8
Qwen3-32B2026.04		8	74	55.2	64.6
Qwen3-4B-Instruct-25072026.04		6.1	67.6	7.6	37.6
Qwen3-8B2026.04		5.4	69.8	4	36.9
Qwen3-8B2026.04		5.3	60.8	35.6	48.2
Qwen3-4B-Thinking-25072026.04		2.7	57.9	19	38.4
GPT-5 mini2026.04		2.4	69.8	48.9	59.3
Qwen3-1.7B2026.04		1.7	50.9	19.3	35.1
Qwen3-1.7B2026.04		0.7	44.9	19.7	32.3