Research map / Large language models / Language, safety and interpretability
Inference-Aware Prompt Optimization for Aligning Black-Box Large Language Models
Saaduddin Mahmud, Mason Nakamura, Kyle Hollins Wray, Shlomo Zilberstein
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| jointly optimize prompt and inference scale (IAPO) instead of inference-agnostic prompt optimization | black-box LLM prompt optimization · six environments | average cost-adjusted reward: up to 25% over disjoint, up to 50% over prompt-only | Improved |
| restrict Top-K screening to K=1 (near-decoupled prompt selection) (ablation) | PSST · Summarization | average cost-adjusted reward | Hurt |
| restrict Top-K screening to K=1 (ablation) | PSST · CommonsenseQA | average cost-adjusted reward | No clear effect |
| add Top-K screening to PSST under small training budget (ablation) | PSST · real-data tasks (MATH, CommonsenseQA, Helpful-Harmless, Summarization) | average cost-adjusted reward | Mixed |
| use PSST structure-aware arm elimination instead of UCB, softmax, epsilon-greedy, uniform | IAPO policy learning · six environments | average cost-adjusted reward | Improved |
| randomly assign N per query rather than optimizing N per context (ablation) | TRIPLE · six environments | average cost-adjusted reward | Hurt |
| choose prompt by single-shot accuracy rather than under majority voting | prompt selection · MATH, MV | accuracy at higher budget | Hurt |
Closest papers by meaning
- Align-Pro: A Principled Approach to Prompt Optimization for LLM Alignment (AAAI 2025 · similarity 0.90)
- Unleashing the Potential of Large Language Models as Prompt Optimizers: Analogical Analysis with Gradient-based Model Optimizers (AAAI 2025 · similarity 0.90)
- GReaTer: Gradients Over Reasoning Makes Smaller Language Models Strong Prompt Optimizers (ICLR 2025 · similarity 0.90)
- Scaling Prompt Synthesis for Large Language Model Reasoning (ICML 2026 · similarity 0.89)
- PRESTO: Preimage-Informed Instruction Optimization for Prompting Black-Box LLMs (NeurIPS 2025 · similarity 0.89)
- Dialogue for Prompting: A Policy-Gradient-Based Discrete Prompt Generation for Few-Shot Learning (AAAI 2024 · similarity 0.89)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
