A Unified Multi-Task Learning Framework for Generative Auto-Bidding with Validation-Aligned Optimization

2025-10-08

AI Summary

In online advertising, the proliferation of highly customized bidding tasks incurs substantial computational overhead and poor data efficiency in multi-task optimization; moreover, existing training-dynamics-based multi-task methods suffer from poor generalization under high-volatility bidding environments. To address these challenges, we propose a unified multi-task learning framework featuring three core innovations: (1) a verification-aligned optimization mechanism that ensures training gradients are consistent with deployment objectives; (2) a period-aware temporal module to enhance cross-task transfer of seasonal patterns; and (3) a gradient-alignment-driven adaptive task weighting strategy. Integrated with generative automated bidding modeling, the framework significantly improves both performance and convergence stability in both simulated environments and large-scale production ad systems, demonstrating strong efficacy and practical applicability.

Technology Category

• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Search and Optimization: Learning to Search
• Multiagent Systems: Adversarial Agents

Application Category

• User Modeling, Personalization and Recommendation: User modeling for targeted and personalized online advertising
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Advertising auctions, pricing, markets, and exchanges

Abstract

In online advertising, heterogeneous advertiser requirements give rise to numerous customized bidding tasks that are typically optimized independently, resulting in extensive computation and limited data efficiency. Multi-task learning offers a principled framework to train these tasks jointly through shared representations. However, existing multi-task optimization strategies are primarily guided by training dynamics and often generalize poorly in volatile bidding environments. To this end, we present Validation-Aligned Multi-task Optimization (VAMO), which adaptively assigns task weights based on the alignment between per-task training gradients and a held-out validation gradient, thereby steering updates toward validation improvement and better matching deployment objectives. We further equip the framework with a periodicity-aware temporal module and couple it with an advanced generative auto-bidding backbone to enhance cross-task transfer of seasonal structure and strengthen bidding performance. Meanwhile, we provide theoretical insights into the proposed method, e.g., convergence guarantee and alignment analysis. Extensive experiments on both simulated and large-scale real-world advertising systems consistently demonstrate significant improvements over typical baselines, illuminating the effectiveness of the proposed approach.

Problem

• Optimizes multiple bidding tasks jointly through shared representations
• Aligns training with validation objectives for better generalization
• Enhances cross-task transfer of temporal patterns in bidding

Innovation

• Validation-aligned multi-task optimization for auto-bidding
• Periodicity-aware temporal module enhances cross-task transfer
• Generative auto-bidding backbone strengthens bidding performance