---
sourceDocument: Yokohama Enable AI
sourceDocumentLink: https://servicenow-prod.fluidtopics.net/r/yokohama/intelligent-experiences

 Release :

    - yokohama

ft:locale :

    - en-US

ft:publication_title :

    - Yokohama Enable AI

ft:clusterId :

    - platai

bundleId :

    - platai

workflow :

    - Platform


---

# Understanding LLMs

# Understanding large language models (LLMs) {#ariaid-title1}

Release version: Yokohama  
Updated November 20, 2025  
![](https://www.servicenow.com/docs/portal-asset/ico-clock) 2 minutes to read
Summarize  
![AI sparkle icon](https://servicenow.com/docs/portal-asset/ai-sparkle-icon) Summarized using AI  
This content was generated using new OpenAI-powered functionality. Results are provided on an as is basis and are not guaranteed to be accurate or complete.  

## Summary of Understanding large language models (LLMs)

Large language models (LLMs) such as ChatGPT and Copilot are advanced AI systems that generate human-like language dynamically.
Unlike retrieval-based systems, LLMs do not store fixed answers but create responses probabilistically, one word at a time.
This generative nature means outputs vary even for the same input, enabling flexibility, creativity, and adaptability.
Show full answer Show less  

## How LLMs Work

LLMs build statistical models from massive text datasets to predict the most probable next word or token in a sequence. This prediction is non-deterministic, so responses differ based on probabilistic sampling rather than fixed scripts. This design allows LLMs to produce diverse and context-sensitive answers.

## Why Results May Vary

* **Probabilistic Sampling:** The model samples among several likely word options instead of always choosing the single most likely one, introducing variability.
* **Temperature Settings:** An internal parameter controlling randomness; higher temperature yields more creative responses, while lower temperature makes outputs more repetitive.
* **Multiple Valid Answers:** Since many questions have multiple correct explanations, the model may phrase responses differently each time.
* **Context Sensitivity:** Minor changes in punctuation or prior conversation context can alter the resulting output.
* **System-Level Factors:** Hardware concurrency, floating-point calculations, or backend updates can cause slight differences in results even with identical inputs.

These intentional variations enhance the LLM's ability to be creative and avoid rigid repetition.

## ServiceNow AI Platform Considerations

The ServiceNow AI Platform supports various LLM models and search tools. Customers should expect that different AI search features may return differing answers for similar queries. This behavior is normal and reflects underlying model variability and search method differences.

Understanding these aspects helps ServiceNow customers effectively use LLM-powered features with realistic expectations about response variability, enabling more flexible and intelligent interactions.  
Large language models are generative, not retrieval-based. They create responses dynamically using probability, which means you can't expect identical outputs every time. This variability is a feature, not a bug, because it
allows for flexibility, creativity, and adaptability.

## How LLMs work {#sn-ai-understanding-llms__section_jqc_55j_jhc}

Large language models (LLMs), like ChatGPT or Copilot, are advanced AI systems that are trained on massive amounts of text to understand and generate human-like language. They build a statistical model of language, so they don't store fixed answers like an
encyclopedia. When you ask a question, the model generates an answer one word (or token) at a time, choosing the next most likely word based on probabilities learned during training. This prediction process makes them powerful, but it's
also why they are non-deterministic. This means the system does not always produce the exact same result (output) for the same prompt (input).

## Why results may vary {#sn-ai-understanding-llms__section_s1j_dwj_jhc}

Even if you provide the same question or prompt twice, the response can differ. Here's why:

Probabilistic sampling
:   The model doesn't always pick the single most likely word. It samples from several likely options. This introduces variation.

Temperature settings
:   Temperature controls randomness, and this internal parameter varies among LLM models. A higher temperature delivers more creative responses, while lower temperatures tend to be more repetitive.

Multiple valid answers
:   Many questions have more than one correct way to explain something. The model may choose different phrasing or emphasis each time.

Context sensitivity
:   Tiny changes in punctuation or prior conversation can shift the output.

System-level factors
:   Hardware concurrency, floating-point math, and backend updates can introduce slight variations, even when everything else is fixed.

For example, think of it like rolling dice to pick words. When you ask a question, the model doesn't follow a fixed script. Instead, it looks at many possible next words and picks one based on probabilities---like rolling weighted dice.
The dice are weighted toward the most likely words, but there's still a chance for variation. If you roll again (ask the same question), you might get a slightly different sequence, even though the rules didn't change. This randomness
is intentional. It makes the model flexible and creative, rather than rigid and repetitive.

For more information about supported LLM models, see [Large language models on the ServiceNow AI Platform](https://servicenow-prod.fluidtopics.net/bLpLFP3BI0aVaJkVBudHHg "ServiceNow AI Platform utilises large language models (LLM) for generative AI and agentic AI functionality. You can choose the ServiceNow LLM or supported third-party models. LLMs bring about the Generative AI application capabilities by understanding and generating human language, and processing vast amount of data.").

## Variations in search {#sn-ai-understanding-llms__section_yvz_nvj_jhc}

The ServiceNow AI Platform® offers a variety of search tools, which may return different answers for the same or similar searches. This disparity in results is expected. For more information, see the following topics:

* [Discrepancies when using different AI search tools](https://servicenow-prod.fluidtopics.net/xLKc9Oik1M3DKT4m6Fus3A "Different AI search tools may return different answers for the same or similar searches. This difference in results is expected. It occurs because each large language model (LLM) uses a different approach to find results and generate answers that match your search.")
* [Search result disparities between AI Search and Now Assist search features](https://www.servicenow.com/docs/access?context=search-disparities-ai-search-now-assist&version=yokohama&pubname=yokohama-platform-administration&ft:locale=en-US)
{#sn-ai-understanding-llms__ul_rjd_dck_jhc}