PersonaTree is a new hierarchical memory framework for persistent LLM agents that structures evidence into persona claims via support paths and outperforms baselines on six person-understanding benchmarks.
Accepted to COLM
7 Pith papers cite this work. Polarity classification is still indexing.
7
Pith papers citing it
citation-role summary
background 3
citation-polarity summary
years
2026 7roles
background 3polarities
background 3representative citing papers
CBEA with LCV bounds evidence sets and validates commitments before response generation, achieving zero failures in scoped tests at 0.49-0.60 availability versus near-zero for baselines.
Omni-Persona benchmark with 18 tasks shows open-source models have audio-visual grounding gaps, RLVR narrows them but leads to conservative outputs, and scale or recall alone fail as diagnostics.
Introduces MPT benchmark and PRefine method that models user preferences as evolving hypotheses to improve personalized tool calling accuracy with 1.24% of full-history token cost.
Reifying ad-hoc preference phrases as GUI widgets, steered by log-probability modulation at decode time, improves precision and perceived control of LLM writing over natural-language prompting alone.
MAESTRO adds a shared preference memory plus GUI-adaptation and workflow-navigation mechanisms to conversational agents with GUIs and tests them in a 33-person movie-booking study.
Debiasing-DPO reduces bias to spurious social contexts by 84% and improves predictive accuracy by 52% on average for LLMs evaluating U.S. classroom transcripts.
citing papers explorer
-
PersonaTree: Structured Lifecycle Memory for Person Understanding in LLM Agents
PersonaTree is a new hierarchical memory framework for persistent LLM agents that structures evidence into persona claims via support paths and outperforms baselines on six person-understanding benchmarks.
-
Recall Isn't Enough: Bounding Commitments in Personalized Language Systems
CBEA with LCV bounds evidence sets and validates commitments before response generation, achieving zero failures in scoped tests at 0.49-0.60 availability versus near-zero for baselines.
-
Omni-Persona: Systematic Benchmarking and Improving Omnimodal Personalization
Omni-Persona benchmark with 18 tasks shows open-source models have audio-visual grounding gaps, RLVR narrows them but leads to conservative outputs, and scale or recall alone fail as diagnostics.
-
Latent Preference Modeling for Cross-Session Personalized Tool Calling
Introduces MPT benchmark and PRefine method that models user preferences as evolving hypotheses to improve personalized tool calling accuracy with 1.24% of full-history token cost.
-
From Words to Widgets for Controllable LLM Generation
Reifying ad-hoc preference phrases as GUI widgets, steered by log-probability modulation at decode time, improves precision and perceived control of LLM writing over natural-language prompting alone.
-
MAESTRO: Adapting GUIs and Guiding Navigation with User Preferences in Conversational Agents with GUIs
MAESTRO adds a shared preference memory plus GUI-adaptation and workflow-navigation mechanisms to conversational agents with GUIs and tests them in a 33-person movie-booking study.
-
Mitigating LLM biases toward spurious social contexts using direct preference optimization
Debiasing-DPO reduces bias to spurious social contexts by 84% and improves predictive accuracy by 52% on average for LLMs evaluating U.S. classroom transcripts.