Factored Reasoning with Inner Speech and Persistent Memory for Evidence-Grounded Human–Robot Interaction Journal: Robotics and Autonomous Systems
Valerio Belcamino, Mariya Kilina, Alessandro Carfì, Valeria Seidita, Fulvio Mastrogiovanni, Antonio Chella · 2026
Read Factored Reasoning with Inner Speech and Persistent Memory for Evidence-Grounded Human–Robot Interaction Journal: Robotics and Autonomous Systems by Valerio Belcamino, Mariya Kilina, Alessandro Carfì, Valeria Seidita, Fulvio Mastrogiovanni, Antonio Chella (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-00675.txt.
Description
Valerio Belcamino and colleagues introduce JANUS, an architecture for assistive robots that must hold user context, recover from vague requests, and ground answers in evidence. Interaction is modelled as a partially observable Markov decision process, realised as a factored controller with typed interfaces and modules for scope, intent, memory, inner speech, queries, and outer speech. Its memory keeps a bounded recent buffer, a compact core store, and an archival store searched by meaning. Inner speech checks parameters and prompts clarification before grounding, while a faithfulness rule ties claims to evidence.
Catalog record
| CREATOR | Valerio Belcamino, Mariya Kilina, Alessandro Carfì, Valeria Seidita, Fulvio Mastrogiovanni, Antonio Chella |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 100150 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2602.00675