Share your thoughts, 1 month free Claude Pro on usSee more

Web navigation on WebShop

76Success Rate

ICRL

Updated 2d ago

Evaluation Results

Method	Links
ICRL 2026.05		76	88.3	-
ICRL 2026.05		74.5	88.9	-
Critique-GRPO 2026.05		73.5	86.8	-
ScalingInter 2026.05		73	86.2	-
Critique-GRPO 2026.05		72.5	87.1	-
GRPO 2026.05		71.5	83.7	-
GSPO 2026.05		71	83.4	-
MATPO 2026.05		68	81.6	-
ScalingInter 2026.05		66.5	84.7	-
GSPO 2026.05		65	82.1	-
MATPO 2026.05		65	82.4	-
GRPO 2026.05		63.5	81.8	-
Gemini-3-Flash 2026.05		53	57.3	-
Gemini-2.5-flash 2026.05		41.2	48.7	-
Qwen2.5-VL-7B 2026.05		18.3	-	8.5
GPT-4o 2026.05		17.2	-	15.5
SCALE 2026.05		14.4	-	5.7
InternVL2.5-8B 2026.05		11.6	-	6.6
Qwen2.5-VL-7B 2026.05		11.2	-	12.7
SCALE 2026.05		11	-	4.9
InternVL2.5-8B 2026.05		8.8	-	16.5
ViGoRL 2026.05		5.6	-	13.4
Augvis 2026.05		4.3	-	-
Qwen2.5-VL-7B 2026.05		4.1	-	19.4
Qwen2.5-VL-7B 2026.05		4.1	-	-
InternVL2.5-8B 2026.05		3.9	-	23.5
Qwen3-30B-A3B 2026.05		3	25.6	-
InternVL2.5-8B 2026.05		3	-	-
SCALE-20k 2026.05		2.1	-	10.1
Qwen3-8B 2026.05		2	10.3	-
Qwen3-4B 2026.05		1.5	9.6	-
LLaVA-NeXT-8B 2026.05		0	-	-
SFT 2025.11		-	63.9	40.3
ETO 2025.11		-	65.7	59.5
Co-Evolving Agents 2025.11		-	72.5	66.3
GPT-4 2025.12		-	-	63.2
GPT-3.5-Turbo 2025.12		-	-	62.4
Llama-2-7B 2025.12		-	-	17.9
SFT 2025.12		-	-	60.2
RFT-PPO 2025.12		-	-	64.2
RFT-CR 2025.12		-	-	63.6
ETO 2025.12		-	-	67.4
Step-PPO 2025.12		-	-	64
IPR 2025.12		-	-	71.3
MACLA 2025.12		-	-	70.2