LaMDA: Language Models for Dialog Applications

Abstract

We present LaMDA: Language Models for Dialog Applications. LaMDA is a family of Transformer-based neural language models specialized for dialog, which have up to 137B parameters and are pre-trained on 1.56T words of public dialog data and web text. While model scaling alone can improve quality, it shows less improvements on safety and factual grounding. We demonstrate that fine-tuning with annotated data and enabling the model to consult external knowledge sources can lead to significant improvements towards the two key challenges of safety and factual grounding. The first challenge, safety, involves ensuring that the model's responses are consistent with a set of human values, such as preventing harmful suggestions and unfair bias. We quantify safety using a metric based on an illustrative set of human values, and we find that filtering candidate responses using a LaMDA classifier fine-tuned with a small amount of crowdworker-annotated data offers a promising approach to improving model safety. The second challenge, factual grounding, involves enabling the model to consult external knowledge sources, such as an information retrieval system, a language translator, and a calculator. We quantify factuality using a groundedness metric, and we find that our approach enables the model to generate responses grounded in known sources, rather than responses that merely sound plausible. Finally, we explore the use of LaMDA in the domains of education and content recommendations, and analyze their helpfulness and role consistency.

Cited by in corpus (58)

• Unifying Large Language Models and Knowledge Graphs: A Roadmap
• Recommender Systems in the Era of Large Language Models (LLMs)
• Compute Trends Across Three Eras of Machine Learning
• The Programmer's Assistant: Conversational Interaction with a Large Language Model for Software Development
• Large AI Models in Health Informatics: Applications, Challenges, and the Future
• Transforming Sentiment Analysis in the Financial Domain with ChatGPT
• Large Language Models and the Reverse Turing Test
• Evaluating Large Language Models on a Highly-specialized Topic, Radiation Oncology Physics
• "What It Wants Me To Say": Bridging the Abstraction Gap Between End-User Programmers and Code-Generating Large Language Models
• Tool Learning with Large Language Models: A Survey
• Theory of Mind for Multi-Agent Collaboration via Large Language Models
• Astronomia ex machina: a history, primer, and outlook on neural networks in astronomy
• Assessing Student Errors in Experimentation Using Artificial Intelligence and Large Language Models: A Comparative Study with Human Raters
• LaMPost: Design and Evaluation of an AI-assisted Email Writing Prototype for Adults with Dyslexia
• Learning from models beyond fine-tuning
• Empowering Molecule Discovery for Molecule-Caption Translation with Large Language Models: A ChatGPT Perspective
• LINC: A Neurosymbolic Approach for Logical Reasoning by Combining Language Models with First-Order Logic Provers
• Reinforcement Learning for Generative AI: State of the Art, Opportunities and Open Research Challenges
• Generative Artificial Intelligence Meets Synthetic Aperture Radar: A Survey
• A Survey on Symbolic Knowledge Distillation of Large Language Models
• Enhanced Automated Code Vulnerability Repair using Large Language Models
• Quo Vadis ChatGPT? From Large Language Models to Large Knowledge Models
• Examining Inter-Consistency of Large Language Models Collaboration: An In-depth Analysis via Debate
• Prompted LLMs as Chatbot Modules for Long Open-domain Conversation
• A Mixed-Methods Approach to Understanding User Trust after Voice Assistant Failures
• Two Directions for Clinical Data Generation with Large Language Models: Data-to-Label and Label-to-Data
• Higher education assessment practice in the era of generative AI tools
• Towards The Ultimate Brain: Exploring Scientific Discovery with ChatGPT AI
• A Framework for Neurosymbolic Robot Action Planning using Large Language Models
• A Multi-Expert Large Language Model Architecture for Verilog Code Generation
• Learning to Automate Follow-up Question Generation using Process Knowledge for Depression Triage on Reddit Posts
• Dungeons and Dragons as a Dialog Challenge for Artificial Intelligence
• A Comprehensive Review of State-of-The-Art Methods for Java Code Generation from Natural Language Text
• ProKnow: Process Knowledge for Safety Constrained and Explainable Question Generation for Mental Health Diagnostic Assistance
• GPT-assisted learning of structure-property relationships by graph neural networks: Application to rare-earth doped phosphors
• Solving Math Word Problems via Cooperative Reasoning induced Language Models
• Artificial intelligence is algorithmic mimicry: why artificial "agents" are not (and won't be) proper agents
• Towards More Human-like AI Communication: A Review of Emergent Communication Research
• Large Language Models to the Rescue: Reducing the Complexity in Scientific Workflow Development Using ChatGPT
• Processing Natural Language on Embedded Devices: How Well Do Transformer Models Perform?
• Reward Guidance for Reinforcement Learning Tasks Based on Large Language Models: The LMGT Framework
• Symbolic Equation Solving via Reinforcement Learning
• Reminding Forgetful Organic Neuromorphic Device Networks
• Improving the Capabilities of Large Language Model Based Marketing Analytics Copilots With Semantic Search And Fine-Tuning
• Blacks is to Anger as Whites is to Joy? Understanding Latent Affective Bias in Large Pre-trained Neural Language Models
• SMILE: Evaluation and Domain Adaptation for Social Media Language Understanding
• DSS: Synthesizing long Digital Ink using Data augmentation, Style encoding and Split generation
• NewsDialogues: Towards Proactive News Grounded Conversation
• InsightNet: Structured Insight Mining from Customer Feedback
• An Overview on Generative AI at Scale with Edge-Cloud Computing
• Towards Improving Interpretability of Language Model Generation through a Structured Knowledge Discovery Approach
• CompanyKG: A Large-Scale Heterogeneous Graph for Company Similarity Quantification
• Tailored-LLaMA: Optimizing Few-Shot Learning in Pruned LLaMA Models with Task-Specific Prompts
• ReFactX: Scalable Reasoning with Reliable Facts via Constrained Generation
• A Hierarchical Neural Framework for Classification and its Explanation in Large Unstructured Legal Documents
• Encapsulating Textual Contents into a MOC data Structure for Advanced Applications
• Emergent inabilities? Inverse scaling over the course of pretraining
• Are Compressed Language Models Less Subgroup Robust?