A gradient-based discrete prompt optimizer that uses reasoning chains to let small LMs self-optimize prompts, outperforming text-feedback baselines on reasoning benchmarks.
In contrast, other methods show markedly less consistency due to their sole reliance on LLM judgment
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
GReaTer: Gradients over Reasoning Makes Smaller Language Models Strong Prompt Optimizers
A gradient-based discrete prompt optimizer that uses reasoning chains to let small LMs self-optimize prompts, outperforming text-feedback baselines on reasoning benchmarks.