Salvato in:
| Autori principali: | , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | https://arxiv.org/abs/2604.25022 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866910172670066688 |
|---|---|
| author | Al-Ratrout, Mohammad Ravva, Pavan Uttej Sharmin, Shayla Raikwar, Aditya Shin, Ju Young Barmaki, Roghayeh Leila |
| author_facet | Al-Ratrout, Mohammad Ravva, Pavan Uttej Sharmin, Shayla Raikwar, Aditya Shin, Ju Young Barmaki, Roghayeh Leila |
| contents | When multiple people share a single voice assistant, the system conflates their histories: one resident's preferences can leak into another's responses, eroding utility and trust. We call this failure mode persona confusion, and we show it is a measurable problem in today's single-user dialogue systems when deployed in shared environments. We present the Adaptive Friend Agent (AFA), a modular framework that combines voice-based speaker identification with per-user memory stores to enable identity-aware, personalized dialogue across multiple users. To support training and evaluation, we construct PAT (Personalized Agent chaT), a synthetic dataset of 58,289 persona-grounded dialogue turns spanning 133 user profiles and 12 real-world scenarios. We evaluate AFA across five LLM back-ends in a standard response-quality benchmark, with a LLaMA-2-70B model fine-tuned on PAT achieving the highest overall performance. To directly measure persona confusion prevention, we introduce an interleaved multi-user evaluation protocol with a novel metric, Persona Attribution Accuracy (PAA), demonstrating that identity-aware routing improves PAA from 35.7% to 61.3%. Human evaluation confirms annotators perceive significantly higher personalization in routing-enabled responses. Our results establish that identity-aware user routing is the critical component for preventing persona confusion in multi-user conversational systems. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_25022 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | AFA: Identity-Aware Memory for Preventing Persona Confusion in Multi-User Dialogue Al-Ratrout, Mohammad Ravva, Pavan Uttej Sharmin, Shayla Raikwar, Aditya Shin, Ju Young Barmaki, Roghayeh Leila Human-Computer Interaction When multiple people share a single voice assistant, the system conflates their histories: one resident's preferences can leak into another's responses, eroding utility and trust. We call this failure mode persona confusion, and we show it is a measurable problem in today's single-user dialogue systems when deployed in shared environments. We present the Adaptive Friend Agent (AFA), a modular framework that combines voice-based speaker identification with per-user memory stores to enable identity-aware, personalized dialogue across multiple users. To support training and evaluation, we construct PAT (Personalized Agent chaT), a synthetic dataset of 58,289 persona-grounded dialogue turns spanning 133 user profiles and 12 real-world scenarios. We evaluate AFA across five LLM back-ends in a standard response-quality benchmark, with a LLaMA-2-70B model fine-tuned on PAT achieving the highest overall performance. To directly measure persona confusion prevention, we introduce an interleaved multi-user evaluation protocol with a novel metric, Persona Attribution Accuracy (PAA), demonstrating that identity-aware routing improves PAA from 35.7% to 61.3%. Human evaluation confirms annotators perceive significantly higher personalization in routing-enabled responses. Our results establish that identity-aware user routing is the critical component for preventing persona confusion in multi-user conversational systems. |
| title | AFA: Identity-Aware Memory for Preventing Persona Confusion in Multi-User Dialogue |
| topic | Human-Computer Interaction |
| url | https://arxiv.org/abs/2604.25022 |