Tesi etd-02052026-155733 |
Link copiato negli appunti
Tipo di tesi
Tesi di laurea magistrale
Autore
BARGIOTTI, LEONARDO
URN
etd-02052026-155733
Titolo
A comparative analysis of neural cognitive models and neural language models
Dipartimento
INGEGNERIA DELL'INFORMAZIONE
Corso di studi
ARTIFICIAL INTELLIGENCE AND DATA ENGINEERING
Relatori
relatore Cimino, Mario Giovanni Cosimo Antonio
relatore Masala, Giovanni L.
relatore Giorgi, loanna
relatore Masala, Giovanni L.
relatore Giorgi, loanna
Parole chiave
- annabell
- artificial intelligence
- cognitive intelligence
- intelligenza cognitiva
- pain
- pleasure
- small llm
Data inizio appello
27/02/2026
Consultabilità
Non consultabile
Data di rilascio
27/02/2096
Riassunto (Inglese)
Riassunto (Italiano)
This thesis investigates how emotional states, specifically pain and pleasure, are represented and inferred within a cognitive architecture, and compares this behavior with that of small-scale large language models. The study focuses on ANNABELL, a biologically inspired system that learns emotional concepts through explicit semantic rules and structured interaction.
ANNABELL is evaluated on narrative scenarios under progressively reduced contextual information and compared with four small language models. Results show that ANNABELL exhibits fully deterministic behavior, maintaining correct inferences as long as the causal structure is preserved, while language models display probabilistic volatility and high sensitivity to minor contextual variations.
ANNABELL is evaluated on narrative scenarios under progressively reduced contextual information and compared with four small language models. Results show that ANNABELL exhibits fully deterministic behavior, maintaining correct inferences as long as the causal structure is preserved, while language models display probabilistic volatility and high sensitivity to minor contextual variations.
File
| Nome file | Dimensione |
|---|---|
La tesi non è consultabile. |
|