Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Open-ended scientific discovery on iNatDisco 800
Loading...
8
Recall
DiscoPER
-0.32
1.84
4
6.16
Jul 1, 2026
Recall
Support Rate
Updated 24d ago
Evaluation Results
Method
Method
Links
Recall
Support Rate
DiscoPER
Type=LLM, H=H_open_cod...
2026.07
8
72.7
DiscoPER w/o Reflect
Type=LLM, H=H_open_cod...
2026.07
7
70
HeurekaBench
Type=LLM, H=H_guided_c...
2026.07
3
62.2
ExperiGen
Type=LLM, H=H_guided_c...
2026.07
3
56.6
GPT-4 BFS
Type=Causal, H=H_edge,...
2026.07
1
-
LLM+PC
Type=Causal, H=H_edge,...
2026.07
0
-
LLM+NOTEARS
Type=Causal, H=H_edge,...
2026.07
0
-
Feedback
Search any
task
Search any
task