List your company
ResearchFormerly WizWand

Interactive Decision Making on ScienceWorld Unseen

85.15Success RateBPO
0.244422.287244.3366.3728Aug 5, 2025Sep 14, 2025Oct 25, 2025Dec 5, 2025Jan 15, 2026Feb 25, 2026Apr 7, 2026
Updated 5mo ago

Evaluation results

MethodLinks
BPOApproach=System-2, Evaluation Protocol=Fine-Tuned, Base Model=Llama-3.1-8B-Instruct2025.0885.15
STEP-HRLBackbone=Llama-3-8B, Evaluation Protocol=Fine-tuning (LoRA)2026.0477.81
MPOApproach=System-2, Evaluation Protocol=Fine-Tuned, Base Model=Llama-3.1-8B-Instruct2025.0877.46
ETOApproach=System-2, Evaluation Protocol=Fine-Tuned, Base Model=Llama-3.1-8B-Instruct2025.0877.25
STEP-HRLBackbone=Mistral-7B, Evaluation Protocol=Fine-tuning (LoRA)2026.0475.21
STEP-HRLBackbone=Gemma-7B, Evaluation Protocol=Fine-tuning (LoRA)2026.0474.08
SFTApproach=System-2, Evaluation Protocol=Fine-Tuned, Base Model=Llama-3.1-8B-Instruct2025.0872.04
GLIDERBackbone=Llama-3-8B, Evaluation Protocol=Fine-tuning (LoRA)2026.0468.34
GLIDERBackbone=Mistral-7B, Evaluation Protocol=Fine-tuning (LoRA)2026.0465.14
Deepseek-R1Approach=System-2, Evaluation Protocol=Zero-shot, Model Category=Reasoning LLMs2025.0859.62
GLIDERBackbone=Gemma-7B, Evaluation Protocol=Fine-tuning (LoRA)2026.0458.5
o3-miniApproach=System-2, Evaluation Protocol=Zero-shot, Model Category=Reasoning LLMs2025.0855.45
WKMBackbone=Llama-3-8B, Evaluation Protocol=Fine-tuning (LoRA)2026.0454.75
WKMBackbone=Mistral-7B, Evaluation Protocol=Fine-tuning (LoRA)2026.0453.62
ETOBackbone=Llama-3-8B, Evaluation Protocol=Fine-tuning (LoRA)2026.0452.33
ETOBackbone=Mistral-7B, Evaluation Protocol=Fine-tuning (LoRA)2026.0451.85
Qwen-3-8B-ThinkingApproach=System-2, Evaluation Protocol=Zero-shot, Model Category=Reasoning LLMs2025.0850.27
WKMBackbone=Gemma-7B, Evaluation Protocol=Fine-tuning (LoRA)2026.0449.24
ETOBackbone=Gemma-7B, Evaluation Protocol=Fine-tuning (LoRA)2026.0447.84
Llama-3.1-8B-InstructApproach=System-1, Evaluation Protocol=Zero-shot, Model Category=Base Models2025.0846.92
SwitchSageBackbone=Mistral-7B, Evaluation Protocol=Prompt-based2026.0445.25
SwitchSageBackbone=Llama-3-8B, Evaluation Protocol=Prompt-based2026.0440.58
REACTBackbone=GPT-4, Evaluation Protocol=Prompt-based2026.0438.05
Qwen-2.5-7B-InstructApproach=System-1, Evaluation Protocol=Zero-shot, Model Category=Base Models2025.0833.18
SwitchSageBackbone=Gemma-7B, Evaluation Protocol=Prompt-based2026.0430.9
ReflexionBackbone=Llama-3-8B, Evaluation Protocol=Prompt-based2026.0425.41
ReActBackbone=Llama-3-8B, Evaluation Protocol=Prompt-based2026.0422.66
ReflexionBackbone=Mistral-7B, Evaluation Protocol=Prompt-based2026.0418.11
ReActBackbone=Mistral-7B, Evaluation Protocol=Prompt-based2026.0417.65
REACTBackbone=GPT-3.5-Turbo, Evaluation Protocol=Prompt-based2026.045.97
ReflexionBackbone=Gemma-7B, Evaluation Protocol=Prompt-based2026.043.93
ReActBackbone=Gemma-7B, Evaluation Protocol=Prompt-based2026.043.51