Related Concept Videos
Stereotype Content Model
Lateralization
Modeling and Similitude
Language and Cognition
Higher Mental Functions of the Brain: Language
Language formation and comprehension take place in the dominant hemisphere. The dominant hemisphere is responsible for understanding the meaning of spoken, written, or sign language, as well as the ability to communicate. For most people, the left hemisphere is the dominant one. The right hemisphere, then, gives tone and emotional context to the...
The Anchoring-and-Adjustment Heuristic
You might also read
Related Articles
Articles linked to this work by shared authors, journal, and citation graph.
Effects of encapsulated algae oil supplements on the production of docosahexaenoic acid-enriched milk in mid-lactation dairy cows.
Differential rumen responses in neonatal ruminants to volatile fatty acids, glucose, and physical stimulation: insights into the drivers of early rumen development.
Related Experiment Video
Updated: Sep 16, 2025

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
Published on: December 6, 2024
Do Language Model Agents Align With Humans in Rating Visualizations? An Empirical Study.
Large language models (LLMs) can simulate human ratings for visualization design tasks, but only when guided by expert confidence. These AI agents complement, but do not replace, traditional user studies.
More Related Videos
07:36Eye Tracking During Visually Situated Language Comprehension: Flexibility and Limitations in Uncovering Visual Context Effects
Published on: November 30, 2018
09:27Using Eye Movements Recorded in the Visual World Paradigm to Explore the Online Processing of Spoken Language
Published on: October 13, 2018
Area of Science:
- Human-computer interaction
- Artificial intelligence
- Data visualization
Background:
- Large language models (LLMs) demonstrate potential in interpreting visual data and design principles.
- The capacity of LLMs to accurately predict human feedback on visualization design remains largely unexplored.
Purpose of the Study:
- To evaluate the alignment between LLM-based agent ratings and human judgments in visualization tasks.
- To investigate the factors influencing the accuracy of LLM predictions for human feedback.
- To explore the utility of LLM agents in rapid prototype evaluation.
Main Methods:
- Three studies were conducted: replication of a human-subject study, simulation of six prior studies using LLM agents, and testing of enhancement techniques.
- LLM agents were employed to provide ratings on visualization tasks, and their performance was compared against human subject data and expert confidence levels.
- Techniques such as input preprocessing and knowledge injection were tested to assess their impact on agent robustness and bias.
Main Results:
- LLM agents showed promising performance in mimicking human-like reasoning and ratings in a replicated study.
- Agent-human alignment in simulated studies correlated positively with experts' pre-experiment confidence.
- Enhancement techniques revealed limitations in robustness and potential for bias in LLM agents.
Conclusions:
- LLM-based agents can effectively simulate human ratings for visualization tasks when guided by high-confidence expert hypotheses.
- LLM agents show potential as a complementary tool for rapid prototype evaluation in visualization design.
- LLM simulations are valuable but cannot substitute for comprehensive user studies.