AI Hackathon
← DEPARTURES
IN1115Open ThemeStatus: Ended

IIIT Bangalore · The Liar's Chatroom

International Institute of Information Technology (IIIT), Bangalore · on Unstop

Cash prize: $264
Free credits: Credits: not stated
Who can enter: Solo OK Country rules: unknown
Code rights: Not checked yet
Submissions closedNov 15

This hackathon has ended. Submissions closed on 2026-11-15 03:30 UTC.

Find the one chat character whose claims clash with the evidence in synthetic group conversations, and climb the leaderboard.

Imported from Unstop and approved by our team. We haven't checked the rules by hand.

One email a week

New events that match your topics, every Monday. Unsubscribe anytime.

Flight details

Registration closed:
2026-11-12 18:29 UTC
Submissions closed:
2026-11-15 03:30 UTC
Event ends:
2026-11-12 18:29 UTC
Cash prize:
$264
Page language:
English

Prizes listed: 1: INR 12,500; 2: INR 7,500; 3: INR 5,000

Project ideas

Evidence-Grounded Contradiction Checker

Flag the one character whose chat statement conflicts with verified evidence.

Architecture: Load custom synthetic training and test conversations → Extract each character's claims and verified evidence facts → Score claim-evidence contradictions for each character → Output the single lying character for CSV submission

Stack: NLP, Machine Learning, Kaggle, CSV submission

Four-Character Liar Ranker

Rank A, B, C, and D from chat turns, then pick the one character whose claim contradicts facts.

Architecture: Parse labeled synthetic group chats into four character turns → Build features from claims, speaker, and verified evidence links → Train a multiclass model to rank the four characters → Select the highest liar score and write the CSV

Stack: NLP, Machine Learning, Kaggle, CSV submission

Semi-Supervised Liar Detector

Use labeled synthetic chats and unlabeled test chats to train a contradiction detector for Kaggle CSV.

Architecture: Load labeled train and unlabeled test conversations → Pseudo-label high-confidence test chats with the current model → Retrain the contradiction model on labeled and pseudo-labeled data → Predict one liar per test chat and write the CSV

Stack: NLP, Machine Learning, Kaggle, CSV submission

FAQ

Can I enter on my own, or do I need a team?
Solo entry is allowed: the event page lists a team size of 1 to 4 members and says you may compete individually or in teams of up to 4.
Who is eligible to take part?
The listing names undergraduate, postgraduate and engineering students plus participants from management, medical, law, and arts, commerce, sciences and other fields, but eligibility is not confirmed here, so read the official rules before registering.

Sources

Open the original event page →

More open AI hackathons