PROMID Lab @FIRE-2025-- Call for Participation
Gautam Kishore Shahi
Hello,
FIRE-2025 PROMID Lab -- Call for Participation (apologies for cross-posting)
We invite you to participate in the 2025 edition of PROMID@FIRE. The PROMID shared task aims to explore methods for identifying misinformation in human and LLM-generated texts, as well as reconstructing the input prompt that likely led to a given piece of LLM-generated misinformative text.
- Subtask 1: Prompt Recovery from LLM generated misinformative text
Given a piece of LLM-generated text with misinformation, participants are tasked to generate a plausible prompt that could have produced the target text.
- Subtask 2: Misinformation Detection in LLM generated text
Given a piece of LLM-generated text with misinformation, the objective is to categorize each datapoint into different categories based on the presence of factual incorrectness in the summaries.
- Subtask 3: Misinformation Detection in Social Media Texts
The objective of this task is to classify tweets related to the 2022 Russo-Ukrainian conflict as misinformation or non-misinformation
Further information: https://promid.github.io/index.html
Register and participate: https://promid.github.io/registration.html.
Important Dates
---------------------
30 July 2025: Registration closes
20 August 2025: End of the evaluation cycle
21 September 2025: Submission of participant papers [CEUR-WS]
15 October 2025: Notification of acceptance for the participant papers [CEUR-WS]
30 October 2025: Camera-ready version of the participant papers due [CEUR-WS]
17-20 December 2025: FIRE Conference (Varanasi, India)