2000 character limit reached
Synergizing In-context Learning with Hints for End-to-end Task-oriented Dialog Systems
Published 24 May 2024 in cs.CL | (2405.15585v3)
Abstract: End-to-end Task-Oriented Dialog (TOD) systems typically require extensive training datasets to perform well. In contrast, LLM based TOD systems can excel even with limited data due to their ability to learn tasks through in-context exemplars. However, these models lack alignment with the style of responses in training data and often generate comprehensive responses, making it difficult for users to grasp the information quickly. In response, we propose SyncTOD that synergizes LLMs with task-specific hints to improve alignment in low-data settings. SyncTOD employs small auxiliary models to provide hints and select exemplars for in-context prompts. With ChatGPT, SyncTOD achieves superior performance compared to LLM-based baselines and SoTA models in low-data settings, while retaining competitive performance in full-data settings.
- A multitask, multilingual, multimodal evaluation of chatgpt on reasoning, hallucination, and interactivity. ArXiv, abs/2302.04023.
- Multiwoz - a large-scale multi-domain wizard-of-oz dataset for task-oriented dialogue modelling. In Conference on Empirical Methods in Natural Language Processing.
- Scaling instruction-finetuned language models. ArXiv, abs/2210.11416.
- Key-value retrieval networks for task-oriented dialogue. ArXiv, abs/1705.05414.
- Google. 2023. Palm 2 technical report. ArXiv, abs/2305.10403.
- Realm: Retrieval-augmented language model pre-training. ArXiv, abs/2002.08909.
- Debertav3: Improving deberta using electra-style pre-training with gradient-disentangled embedding sharing.
- Fg2seq: Effectively encoding knowledge for end-to-end task-oriented dialog. ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pages 8029–8033.
- Task-oriented dialog generation with enhanced entity representation. In Interspeech.
- A simple language model for task-oriented dialogue. ArXiv, abs/2005.00796.
- In-context learning for few-shot dialogue state tracking. In Conference on Empirical Methods in Natural Language Processing.
- Autoregressive entity generation for end-to-end task-oriented dialog. ArXiv, abs/2209.08708.
- Vojtech Hudecek and Ondrej Dusek. 2023. Are llms all you need for task-oriented dialogue? ArXiv, abs/2304.06556.
- Prompting for explanations improves adversarial nli. is this true? {Yes} it is {true} because {it weakens superficial cues}. In Findings.
- Retrieval-augmented generation for knowledge-intensive nlp tasks. ArXiv, abs/2005.11401.
- Guiding large language models via directional stimulus prompting. ArXiv, abs/2302.11520.
- Rensis Likert. 1932. A technique for the measurement of attitude scales.
- Bitod: A bilingual multi-domain dataset for task-oriented dialogue modeling. arXiv preprint arXiv:2106.02787.
- What makes good in-context examples for gpt-3? In Workshop on Knowledge Extraction and Integration for Deep Learning Architectures; Deep Learning Inside Out.
- Ilya Loshchilov and Frank Hutter. 2017. Decoupled weight decay regularization. In International Conference on Learning Representations.
- Learning knowledge bases with parameters for task-oriented dialogue systems. ArXiv, abs/2009.13656.
- Mem2seq: Effectively incorporating knowledge bases into end-to-end task-oriented dialog systems. ArXiv, abs/1804.08217.
- Using in-context learning to improve dialogue safety. ArXiv, abs/2302.00871.
- Rodrigo Nogueira and Kyunghyun Cho. 2019. Passage re-ranking with bert. ArXiv, abs/1901.04085.
- Bleu: a method for automatic evaluation of machine translation. In Annual Meeting of the Association for Computational Linguistics.
- Entity-consistent end-to-end task-oriented dialogue system with kb retriever. ArXiv, abs/1909.06762.
- End-to-end task-oriented dialogue: A survey of tasks, methods, and future directions.
- Dynamic fusion network for multi-domain end-to-end task-oriented dialog. In Annual Meeting of the Association for Computational Linguistics.
- Constraint based knowledge base distillation in end-to-end task oriented dialogs. ArXiv, abs/2109.07396.
- In-context retrieval-augmented language models. Transactions of the Association for Computational Linguistics, 11:1316–1331.
- A network-based end-to-end trainable task-oriented dialogue system. In Conference of the European Chapter of the Association for Computational Linguistics.
- Dialokg: Knowledge-structure aware task-oriented dialogue generation. ArXiv, abs/2204.09149.
- Replug: Retrieval-augmented black-box language models. ArXiv, abs/2301.12652.
- Q-tod: A query-driven task-oriented dialogue system. In Conference on Empirical Methods in Natural Language Processing.
- Llama 2: Open foundation and fine-tuned chat models. ArXiv, abs/2307.09288.
- Multi-grained knowledge retrieval for end-to-end task-oriented dialog. In Annual Meeting of the Association for Computational Linguistics.
- Learning to retrieve in-context examples for large language models. ArXiv, abs/2307.07164.
- Sequence-to-sequence learning for task-oriented dialogue with dialogue state representation. In International Conference on Computational Linguistics.
- Global-to-local memory pointer networks for task-oriented dialogue. ArXiv, abs/1901.04713.
- Graphmemdialog: Optimizing end-to-end task-oriented dialog systems using graph memory networks. In AAAI Conference on Artificial Intelligence.
- C-pack: Packaged resources to advance general chinese embedding.
- Unifiedskg: Unifying and multi-tasking structured knowledge grounding with text-to-text language models. ArXiv, abs/2201.05966.
- Pomdp-based statistical spoken dialog systems: A review. Proceedings of the IEEE, 101:1160–1179.
- Enhancing performance on seen and unseen dialogue scenarios using retrieval-augmented end-to-end task-oriented system. In SIGDIAL Conferences.
- Retrieve anything to augment large language models. ArXiv, abs/2310.07554.
- Chatbot arena: Benchmarking llms in the wild with elo ratings. https://lmsys.org/blog/2023-05-03-arena/.
Paper Prompts
Sign up for free to create and run prompts on this paper using GPT-5.
Top Community Prompts
Collections
Sign up for free to add this paper to one or more collections.