Iterative multi-step prompting with feedback gates is reported to improve SimpleBench scores of base LLMs, but the evidence is underpowered and partly rests on an ad hoc metric.
If the feedback gate identifies issues, the model generates a revised step based on the feedback provided
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2024 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
A NotSo Simple Way to Beat Simple Bench
Iterative multi-step prompting with feedback gates is reported to improve SimpleBench scores of base LLMs, but the evidence is underpowered and partly rests on an ad hoc metric.