Challenge: Recent self-learning methods based on user satisfaction metrics and contextual bandits have shown promising results to enable consistent improvements in conversational AI systems.
Approach: They propose a meta-gradient learning approach that adjusts constraint violation penalty terms adaptively through a user-defined meta objective that encourages balanced constraint satisfaction across domains.
Outcome: The proposed framework supports fine-grained exploration targets for individual domains via user-defined constraints.

Similar Papers

Scalable and Robust Self-Learning for Skill Routing in Large-Scale Conversational AI Systems (2022.naacl-industry)

Copied to clipboard

Challenge: Existing methods to enable skill routing do not scale in terms of the number of skills and skill on-boarding.
Approach: They propose a model-based approach to enable natural conversation by allowing frequent policy updates . they propose an annotation-based system, rule-based model, and bandit-based learning .
Outcome: The proposed method is scalable and cost-effective, the authors show . they show that it can improve the user experience without abrupt policy changes .
One Planner To Guide Them All ! Learning Adaptive Conversational Planners for Goal-oriented Dialogues (2025.emnlp-main)

Copied to clipboard

Challenge: Existing methods for goal-oriented dialogues involve training separate models for specific combinations of objectives, leading to computational and scalability issues.
Approach: They propose a new dialogue policy method that can adapt to varying objective preferences at inference time without retraining.
Outcome: The proposed method can adapt to varying objective preferences at inference time without retraining.
Few-Shot Structured Policy Learning for Multi-Domain and Multi-Task Dialogues (2023.findings-eacl)

Copied to clipboard

Challenge: Reinforcement learning is widely adopted to model dialogue managers in task-oriented dialogues, but the user simulator provided by state-of-the-art dialogue frameworks are only rough approximations of human behaviour.
Approach: They propose to use structured policies to improve sample efficiency when learning on multi-domain and multi-task environments.
Outcome: The proposed policies improve sample efficiency and performance on multi-domain and multi-task environments.
Self-Aware Feedback-Based Self-Learning in Large-Scale Conversational AI (2022.naacl-industry)

Copied to clipboard

Challenge: Large-scale conversational AI systems require constant update to adapt to changing customer behavior and trends . lack of self-awareness in feedback-based systems can cause degradation of performance . et al., e. alderman and scott k. d. argues that such systems are not scalable enough to sustain the rapid update pace of conversational systems.
Approach: They propose a superposition-based model that reactively learns local-adaptive decision boundaries . they propose rewritings with a bi-variate beta setting to improve the model's performance .
Outcome: The proposed model improves the PR-AUC by 27.45% and reduces relative defect reductions by 31.22% . the proposed model can adapt faster to changes in global preferences across a large number of customers .
Guided Dialog Policy Learning: Reward Estimation for Multi-Domain Task-Oriented Dialog (D19-1)

Copied to clipboard

Challenge: Existing methods to learn dialog policy require elaborate design and user goals.
Approach: They propose an algorithm that estimates the reward signal and infers the user goal in dialog sessions.
Outcome: The proposed algorithm achieves higher task success than state-of-the-art models on a multi-domain task-oriented dialog dataset.
DyBBT: Dynamic Balance via Bandit-inspired Targeting for Dialog Policy with Cognitive Dual Systems (2026.acl-long)

Copied to clipboard

Challenge: Task oriented dialog systems often rely on static exploration strategies that do not adapt to dynamic dialog contexts.
Approach: They propose a dialog policy learning framework that formalizes the exploration challenge through a structured cognitive state space C.
Outcome: The proposed framework achieves SOTA performance in success rate, efficiency, and generalization.
Conversational Graph Grounded Policy Learning for Open-Domain Conversation Generation (2020.acl-main)

Copied to clipboard

Challenge: Existing word-level policy models that learn dialog policy and language generation from dialog corpora often lead to degeneration issues where the utterances become ungrammatical or repetitive.
Approach: They propose to represent prior dialog transitions as a graph and learn a CG grounded dialog policy that can foster a more coherent and controllable dialog.
Outcome: The proposed framework is able to learn dialog policy in open-domain multi-turn conversation.
Target-Guided Open-Domain Conversation (P19-1)

Copied to clipboard

Challenge: a new study aims to improve opendomain chat systems by integrating goals and strategy into the system.
Approach: They propose a structured approach that introduces coarse-grained keywords to control intended content of system responses and attains smooth conversation transition through turn-level supervised learning.
Outcome: The proposed system produces meaningful and effective conversations significantly better than other approaches.
Gradient-Adaptive Policy Optimization: Towards Multi-Objective Alignment of Large Language Models (2025.acl-long)

Copied to clipboard

Challenge: Reinforcement learning from human feedback (RLHF) has emerged as a powerful technique for aligning large language models (LLMs) with human preferences.
Approach: They propose a novel algorithm that uses multiple-gradient descent to optimize LLMs with diverse preferences to maximize trade-offs between objectives.
Outcome: The proposed approach incorporates user preferences across different objectives and achieves Pareto solutions that better align with the user’s specific needs.
Rethinking Supervised Learning and Reinforcement Learning in Task-Oriented Dialogue Systems (2020.findings-emnlp)

Copied to clipboard

Challenge: Dialogue policy learning for task-oriented dialogue systems has enjoyed great progress through using reinforcement learning methods.
Approach: They propose a dialogue action decoder and a simulator-free adversarial learning method to improve dialogue agent performance without using reinforcement learning.
Outcome: The proposed methods achieve more stable and higher performance with fewer efforts, such as the domain knowledge required to design a user simulator and the intractable parameter tuning in reinforcement learning.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations