LLM-OSDA: An Optimal-Stopping Dynamic Auction for Native Advertising in Multi-Turn LLM Conversations

• URL: http://arxiv.org/abs/2608.00123v2
• Date: Tue, 04 Aug 2026 07:53:57 GMT
• Title: LLM-OSDA: An Optimal-Stopping Dynamic Auction for Native Advertising in Multi-Turn LLM Conversations
• Abstract summary: We propose a dynamic cost-per-click auction that integrates Bellman optimal stopping, winner allocation, and envelope pricing.<n>Experiments on a simulated conversational advertising corpus show that LLM-OSDA improves net revenue by 11 percent over the strongest fixed-timing baseline.
• Score: 27.829715473956472
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: LLM-native advertising embeds sponsored content directly into model-generated responses, shifting the unit of sale from a fixed slot to a moment within an evolving conversation. Existing LLM ad-auction mechanisms primarily operate within a single response, settling the winner but not the timing. The extension is nontrivial: with one native insertion opportunity per session, the stopping time depends on bids, coupling timing with allocation, so static truthfulness arguments no longer apply. We propose the LLM-based Optimal Stopping Dynamic Auction (LLM-OSDA), a dynamic cost-per-click auction that integrates Bellman optimal stopping, winner allocation, and envelope pricing. A bid-independent LLM layer estimates contextual click quality and seamlessly renders the winning ad, while bids enter only the committed auction mechanism. Under an exact Bellman oracle, the expected discounted-click allocation is monotone in each advertiser's bid, and the corresponding envelope payment makes truthful bidding weakly dominant in expectation. For practical deployment, a learned StopNet approximates the Bellman action values. We show that its decisions differ from the optimal policy only near the stopping boundary and bound the resulting incentive loss in terms of its approximation error. Experiments on a simulated conversational advertising corpus show that LLM-OSDA improves net revenue by 11 percent over the strongest fixed-timing baseline while maintaining comparable user retention. Code is at https://github.com/2025Fang2025/llm-osda.

Related papers

• LERA: LLM-Enhanced RAG for Ad Auction in Generative Chatbots [38.35931897144688]
  We propose LERA, a two-stage retrieve-then-generate auction framework for large language model (LLM)-based chatbots.<n>In the first stage, embedding-based coarse filtering pre-selects a small set of candidate advertisers.<n>In the second stage, the LLM itself is queried with a carefully designed prompt to produce logits over candidates, which serve as refined organic relevance scores.
  arXiv  Detail & Related papers  (2026-05-15T15:39:30Z)
• LLM Advertisement based on Neuron Auctions [16.15577957062828]
  generative advertising emerges as a crucial monetization strategy.<n>Existing methods, such as prompt injection or rigid position slots, disrupt semantic coherence.<n>We introduce Neuron Auctions, a novel paradigm that shifts the auction object from the surface text space to the LLM's internal representations.
  arXiv  Detail & Related papers  (2026-05-08T16:54:39Z)
• Anchor-and-Resume Concession Under Dynamic Pricing for LLM-Augmented Freight Negotiation [5.225967186005105]
  Deriving $$ from the live spread enables adaptation but introduces a new problem: a pricing shift can cause the formula to retract a previous offer, violating monotonicity.<n>We propose a two-index anchor-and-resume framework that addresses both limitations.<n>A spread-derived $$ maps each load's margin structure to the correct concession posture, while the anchor-and-resume mechanism guarantees monotonically nondecreasing offers under arbitrary pricing shifts.
  arXiv  Detail & Related papers  (2026-04-22T16:17:07Z)
• $\nabla$-Reasoner: LLM Reasoning via Test-Time Gradient Descent in Latent Space [71.23672814629448]
  $nabla$-Reasoner is an iterative generation framework that integrates differentiable optimization over token logits into the decoding loop.<n>$nabla$-Reasoner achieves over 20% accuracy improvement on a challenging mathematical reasoning benchmark.
  arXiv  Detail & Related papers  (2026-03-05T08:42:54Z)
• LLM-Auction: Generative Auction towards LLM-Native Advertising [10.695066036409274]
  We propose a learning-based generative auction mechanism that integrates auction and LLM generation for LLM-native advertising.<n>We introduce Iterative Reward-Preference Optimization (IRPO) algorithm that alternately optimize the reward model and the LLM.<n>We show that LLM-Auction significantly outperforms existing baselines in allocation efficiency, while achieving the desired mechanism properties.
  arXiv  Detail & Related papers  (2025-12-11T11:31:20Z)
• Reasoning-Aligned Perception Decoupling for Scalable Multi-modal Reasoning [95.44766931218896]
  Multi-modal large language models (MLLMs) still lag behind text-based reasoning.<n>We introduce Perception-Reasoning Decoupling, which modularizes the MLLM's reasoning component and makes it easily replaceable.<n>We propose a novel reinforcement learning algorithm called Visual Perception Optimization (VPO) to align the MLLM's perceptual output with the final reasoning task.
  arXiv  Detail & Related papers  (2025-06-05T02:28:07Z)
• LLM-Lasso: A Robust Framework for Domain-Informed Feature Selection and Regularization [59.75242204923353]
  We introduce LLM-Lasso, a framework that leverages large language models (LLMs) to guide feature selection in Lasso regression.<n>LLMs generate penalty factors for each feature, which are converted into weights for the Lasso penalty using a simple, tunable model.<n>Features identified as more relevant by the LLM receive lower penalties, increasing their likelihood of being retained in the final model.
  arXiv  Detail & Related papers  (2025-02-15T02:55:22Z)
• Procurement Auctions via Approximately Optimal Submodular Optimization [53.93943270902349]
  We study procurement auctions, where an auctioneer seeks to acquire services from strategic sellers with private costs. Our goal is to design computationally efficient auctions that maximize the difference between the quality of the acquired services and the total cost of the sellers.
  arXiv  Detail & Related papers  (2024-11-20T18:06:55Z)
• Dynamic Uncertainty Ranking: Enhancing Retrieval-Augmented In-Context Learning for Long-Tail Knowledge in LLMs [50.29035873837]
  Large language models (LLMs) can learn vast amounts of knowledge from diverse domains during pre-training.<n>Long-tail knowledge from specialized domains is often scarce and underrepresented, rarely appearing in the models' memorization.<n>We propose a reinforcement learning-based dynamic uncertainty ranking method for ICL that accounts for the varying impact of each retrieved sample on LLM predictions.
  arXiv  Detail & Related papers  (2024-10-31T03:42:17Z)
• Know in AdVance: Linear-Complexity Forecasting of Ad Campaign Performance with Evolving User Interest [31.40023913908308]
  Real-time Bidding (RTB) advertisers wish to textitknow in advance the expected cost and yield of ad campaigns to avoid trial-and-error expenses. We propose textitAdVance, a time-aware framework that integrates local auction-level and global campaign-level modeling. AdVance has been deployed on the Tencent Advertising platform, and A/B tests show a remarkable 4.5% uplift in Average Revenue per User (ARPU)
  arXiv  Detail & Related papers  (2024-05-17T10:22:36Z)
• Autobidders with Budget and ROI Constraints: Efficiency, Regret, and Pacing Dynamics [53.62091043347035]
  We study a game between autobidding algorithms that compete in an online advertising platform.<n>We propose a gradient-based learning algorithm that is guaranteed to satisfy all constraints and achieves vanishing individual regret.
  arXiv  Detail & Related papers  (2023-01-30T21:59:30Z)
• ADEPT: A DEbiasing PrompT Framework [64.54665501064659]
  Finetuning is an applicable approach for debiasing contextualized word embeddings.<n> discrete prompts with semantic meanings have shown to be effective in debiasing tasks.<n>We propose ADEPT, a method to debias PLMs using prompt tuning while maintaining the delicate balance between removing biases and ensuring representation ability.
  arXiv  Detail & Related papers  (2022-11-10T08:41:40Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.