Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Scientific Ideation on NeurIPS 2025 (CV)
Loading...
6.74
Novelty
Agentic-Ideation
5.8976
6.1163
6.335
6.5537
Jun 30, 2026
Novelty
Significance
Feasibility
Overall Score
Updated 24d ago
Evaluation Results
Method
Method
Links
Novelty
Significance
Feasibility
Overall Score
Agentic-Ideation
Backbone=Qwen3-8B
2026.06
6.74
7.65
6.13
6.69
SciPIP
Underlying LLM=GPT-4o
2026.06
6.16
7.03
5.54
6.02
SciMON
Underlying LLM=GPT-4o
2026.06
6.09
6.93
5.62
5.99
Qwen-3-8B
Type=Backbone model
2026.06
6.07
6.89
5.46
5.85
ResearchAgent
Underlying LLM=GPT-4o
2026.06
6.02
6.84
5.61
5.94
VirSci
Underlying LLM=GPT-4o
2026.06
5.93
6.91
5.5
5.94
Feedback
Search any
task
Search any
task