paper-with-me

Papers

Inference-Aware Prompt Optimization for Aligning Black-Box Large Language Models

2025-08-08 · Saaduddin Mahmud, Mason Nakamura, Kyle Hollins Wray, Shlomo Zilberstein arxiv

Prompt optimization methods have demonstrated significant effectiveness in aligning black-box large language models (LLMs). In parallel, inference scaling strategies such as Best-of-N Sampling and Majority Voting have likewise been shown to improve alignment and performance by trading additional computation for better output. However, existing prompt optimization approaches are inference strategy agnostic; that is, they optimize prompts without accounting for the inference strategy. This constitutes a significant methodological gap, as our empirical and theoretical analysis reveals a strong interdependence between these two paradigms. Moreover, we find that user preferences regarding trade-offs among multiple objectives and inference budgets substantially influence the choice of prompt and inference configuration. To address this gap, we introduce a novel unified framework named IAPO (Inference-Aware Prompt Optimization) that jointly optimizes the prompt and inference scale, while being aware of the inference budget and different task objectives. We then develop a fixed-budget training algorithm for IAPO, called PSST (Prompt Scaling via Sequential Trimming), and establish finite-budget guarantees on the error probability. Finally, we evaluate the effectiveness of PSST on six tasks, including multi-objective text generation and reasoning, and demonstrate the critical role of incorporating inference-awareness in aligning black-box LLMs using prompt optimization.

📄 PDF Abstract BibTeX arXiv:2508.10030

Code (0)

등록된 구현이 없습니다.

Tasks

Text Generation

Similar Papers 제목 키워드 기반

Aligning LLMs on a Budget: Inference-Time Alignment with Heuristic Reward Models

2025-08-07 · Mason Nakamura, Saaduddin Mahmud, Kyle H. Wray, Hamed Zamani 외 arxiv

Aligning LLMs with user preferences is crucial for real-world use but often requires costly fine-tuning or expensive inference, forcing trade-offs between alignment quality and computational cost. Existing inference-time…

Sharpness-Aware Black-Box Optimization

2024-10-16 · Feiyang Ye, Yueming Lyu, Xuehao Wang, Masashi Sugiyama 외

Black-box optimization algorithms have been widely used in various machine learning problems, including reinforcement learning and prompt fine-tuning. However, directly optimizing the training loss value, as commonly don…

Black-Box Tuning for Language-Model-as-a-Service

2022-01-10 · Tianxiang Sun, Yunfan Shao, Hong Qian, Xuanjing Huang 외

Extremely large pre-trained language models (PTMs) such as GPT-3 are usually released as a service. It allows users to design task-specific prompts to query the PTMs through some black-box APIs. In such a scenario, which…

In-Context LearningLanguage ModelingLanguage Modelling

Route to Rome Attack: Directing LLM Routers to Expensive Models via Adversarial Suffix Optimization

2026-04-16 · Haochun Tang, Yuliang Yan, Jiahua Lu, Huaxiao Liu 외 arxiv

Cost-aware routing dynamically dispatches user queries to models of varying capability to balance performance and inference cost. However, the routing strategy introduces a new security concern that adversaries may manip…

ContraPrompt: Contrastive Prompt Optimization via Dyadic Reasoning Trace Analysis

2026-04-20 · Rishav Rishav, Pushpak Pujari, Pushpendre Rastogi arxiv

Prompt optimization methods either analyze individual failures in isolation or compare prompt variants across examples, operating on single execution traces with no access to the reasoning process distinguishing success …