This program is tentative and subject to change.
Thu 27 Feb 2025 10:00 - 12:00 at Exhibit Hall C - Posters - Posters #1
Large Language Model (LLM)-based assistance has seen rapid adoption in CS1 courses, promising quick and always-available support for homework help. However, a significant challenge lies in ensuring that the hints and guidance provided by these models are accurate and not based on hallucinated solutions. In this work, we study and categorize LLM behavior in cases where students believe an LLM-powered homework bot has given an inaccurate hint. We then demonstrate how certain student behaviors are correlated with our suggested bot-behavior categories.
This program is tentative and subject to change.
Thu 27 FebDisplayed time zone: Eastern Time (US & Canada) change
Thu 27 Feb
Displayed time zone: Eastern Time (US & Canada) change