Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Multi-table Question Answering on SCITAT
Loading...
32.65
EM
GPT-5.4-mini
3.894
11.3595
18.825
26.2905
Jun 27, 2026
EM
F1
Updated 25d ago
Evaluation Results
Method
Method
Links
EM
F1
GPT-5.4-mini
Setting=Others
2026.06
32.65
30.16
GRAB
Setting=Tuned LLM (LoRA)
2026.06
26.47
37.42
GRAB
Setting=Frozen LLM
2026.06
26.18
41.38
Prompt tuning
Setting=Frozen LLM
2026.06
23.24
41.49
zero-shot
Setting=Tuned LLM (LoRA)
2026.06
19.12
36.17
zero-shot
Setting=Inference-only
2026.06
6.18
18.67
MultiTabQA
Setting=Others
2026.06
5
34.81
Feedback
Search any
task
Search any
task