Skip to content

Commit 0a63563

Browse files
s-sajid-aliclaude
andcommitted
Reduce Tools slide image size to 50%
Adjusted the tool-call.png image on slide 22 (Tools) from 75% to 50% width to make it more appropriately sized. The image maintains its aspect ratio while taking up less space on the slide. Co-Authored-By: Claude Sonnet 4.5 <[email protected]>
1 parent 5b33246 commit 0a63563

File tree

1 file changed

+38
-1
lines changed

1 file changed

+38
-1
lines changed

presentation/slides.md

Lines changed: 38 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -721,6 +721,9 @@ prompt = prompt = """
721721
722722
The `date` field is missing in the prompt, put some LLMs hallucinate one.
723723
724+
What happens if you set that field to be either a `str` or `None`?
725+
726+
724727
:: right ::
725728
726729
<SpeechBubble position="r" color="cyan-light" textAlign="left" shape="round" maxWidth="400px">
@@ -757,6 +760,40 @@ The `date` field is missing in the prompt, put some LLMs hallucinate one.
757760
```
758761
</SpeechBubble>
759762
763+
---
764+
layout: top-title-two-cols
765+
columns: is-8
766+
color: violet-light
767+
hideInToc: true
768+
---
769+
770+
[^ref1]: <div class="ns-c-cite">Adapted from <a href="https://platform.openai.com/docs/guides/function-calling">Function calling, OpenAI</a>.</div>
771+
772+
:: title ::
773+
774+
# Tools
775+
776+
:: left ::
777+
778+
- Until this point we send all the data to the LLM within the prompt and expect an answer that is reasonable.
779+
780+
- However, some queries require additional context that can only be gathered when the LLM is queried.
781+
782+
- We can pass function/tool definitions to an LLM and let it decide if it needs to be called (or require it to be used).
783+
784+
- There are some tools that
785+
786+
787+
:: right ::
788+
789+
<div>
790+
<img src="/tool-call.png" style="width: 50%;"/>
791+
</div>
792+
793+
Tool call schematic [^ref1]
794+
795+
796+
760797
761798
---
762799
layout: side-title
@@ -779,5 +816,5 @@ Reach out to Research Technology Services, https://services.rt.nyu.edu/
779816
We have learnt:
780817
- What LLMs are.
781818
- How they generate text.
782-
- What embeddings are.
819+
- How you can generate structured outputs from LLMs.
783820
- How NYU facilitates your access to AI resources.

0 commit comments

Comments
 (0)