A causal explainable guardrails for large language models
Large Language Models (LLMs) have shown impressive performance in natural language
tasks, but their outputs can exhibit undesirable attributes or biases. Existing methods for …
tasks, but their outputs can exhibit undesirable attributes or biases. Existing methods for …
Continual treatment effect estimation: Challenges and opportunities
A further understanding of cause and effect within observational data is critical across many
domains, such as economics, health care, public policy, web mining, online advertising, and …
domains, such as economics, health care, public policy, web mining, online advertising, and …
Continual Causal Effect Estimation
This chapter discusses the convergence of two fast-growing research areas, ie, causal
inference and continual learning. In particular, we discuss the motivating examples, formally …
inference and continual learning. In particular, we discuss the motivating examples, formally …
Informative and Long-Term Response Generation using Multiple Suggestions and User Persona Retrieval in a Dialogue System
Enhancing user satisfaction in dialogue systems relies on their ability to understand users
and generate responses that meet their expectations. This study proposes a dialogue …
and generate responses that meet their expectations. This study proposes a dialogue …
Position Debiasing Fine-Tuning for Causal Perception in Long-Term Dialogue
The core of the dialogue system is to generate relevant, informative, and human-like
responses based on extensive dialogue history. Recently, dialogue generation domain has …
responses based on extensive dialogue history. Recently, dialogue generation domain has …
Sequential Topic Selection Model with Latent Variable for Topic-Grounded Dialogue
Recently, topic-grounded dialogue system has attracted significant attention due to its
effectiveness in predicting the next topic to yield better responses via the historical context …
effectiveness in predicting the next topic to yield better responses via the historical context …