Frequently asked questions
Common questions about psychological state measurement, the science it rests on, and how AI teams work with Receptiviti Labs.
THE MEASUREMENT
What does Receptiviti measure in human-AI interactions?
Receptiviti measures 200+ validated psychological dimensions derived from language: cognitive load, emotional tone, distress signals, rapport, wellbeing trajectory, over-reliance, analytical thinking, and more. These are explicitly computed scores, not model inference. The same input yields the same score every time, traceable to peer-reviewed science.
How is this different from LLM-based inference about psychological state?
Every AI system forms an internal representation of the person it is talking to, and that representation shapes every response. It exists only inside the model, where it cannot be inspected, reused, or compared across systems. Receptiviti replaces that implicit representation with an explicit, structured measurement computed from validated psycholinguistic frameworks. Every score is deterministic, version-stable across model changes, and traceable to peer-reviewed science.
What is the psychological state measurement layer?
The psychological state measurement layer is the structured, validated, auditable representation of what an AI interaction is doing to the person having it, computed from language and independent of the model being assessed. It includes dimensions such as cognitive load trajectory, emotional state, distress signals, dependency patterns, rapport, and agency. Because it is measured rather than inferred, it can be reused across evaluation, supplied back to the model as context, and kept as a record for governance. This is the layer current AI observability stacks do not have.
THE SCIENCE
What is the scientific basis for Receptiviti's measurement?
Receptiviti's measurement is built on LIWC, Linguistic Inquiry and Word Count, created by Dr. James W. Pennebaker and refined over five major versions spanning 30 years. Receptiviti holds the exclusive commercial rights to LIWC. The psycholinguistic frameworks behind Receptiviti's dimensions are grounded in 34,000+ independent peer-reviewed citations spanning health psychology, personality science, clinical research, and organizational behavior.
Does Receptiviti diagnose mental health conditions?
No. Receptiviti's dimensions are research-validated measures of psychological state expressed in language, not diagnostic instruments. They describe patterns such as cognitive load, emotional tone, and distress signals as they appear in text. They do not identify clinical conditions, and they are not a substitute for clinical assessment. In AI systems, they are used to observe how an interaction is affecting a person and to trigger escalation or handoff where appropriate.
WORKING WITH RECEPTIVITI LABS
Who uses Receptiviti Labs?
Receptiviti Labs works with AI safety and alignment teams at foundation model companies, eval engineers and LLM observability teams, enterprise AI deployment teams, companion and long-horizon AI products, and AI governance and responsible AI programs.
How do teams access the measurement layer?
Through a REST API that returns structured scores in under 65ms, fast enough to run on every conversational turn, or through an on-premises container for privacy-sensitive environments. Output is deterministic JSON that can be appended to existing logs and traces, so one measurement serves evaluation, inference, and governance without being computed three times.
How does Receptiviti Labs relate to Receptiviti?
Receptiviti Labs is the AI-focused arm of Receptiviti, established in 2026. Receptiviti was founded in 2015 to measure psychological state from language, and its measurement science is built on LIWC, created by co-founder Dr. James W. Pennebaker. Receptiviti Labs develops the research, methods, and infrastructure that apply that science to AI systems.