Abstract

Conversational agents, chatbots and large language models : from Eliza to ChatGPT and ChatLLAMA. How have very large conversational models been trained ? What are their limits, ethical issues, uses and future ?

Conversing with the machine

Abstract

Speaker(s)

Benoît Sagot

Events

Representing text units

Some examples of NLP applications in the digital humanities

Symbolic and probabilistic approaches

Two examples of the use of transducers in linguistics

Language templates

Learning a language model from audio

Automatic translation

Massively multilingual neural translation

Neural approaches to some application tasks

Knowledge-based text generation

Computational linguistics

Automatic analysis of argumentation in political debates

Conversing with the machine

Prediction is understanding : a neuro-cognitive model of language based on prediction

Multimodalities : NLP and images, NLP and speech

Goal-oriented AI : towards machines capable of learning, reasoning and planning

See also