AAI Labs
MainServicesCasesResearchNewsBlogPublic R&DCareersContact

Research

Research
2026

Towards Reliable Instruction-Following in LLM Multi-Turn Workflows: An Empirical Comparison of Prompting and Control Architectures

LLMs quietly drift from their instructions as conversations grow longer. We show that dynamic control architectures significantly outperform standard prompting at keeping them on track.

Research
2026

Ablating Retrieval Modules for Temporal Conflict Resolution in Legal RAG

When laws change, RAG pipelines hallucinate outdated answers. We ablate retrieval components to find which combinations keep legal AI systems factually grounded over time.

UAB Taikomasis dirbtinis intelektas © 2026

[email protected]LinkedIn →