mlsanityData to 14 Oct 2026

No Trust Issues Here: A Technical Report on the Winning Solutions for the Rayan AI Contest

arXiv 2512.01498 ↗First posted 1 Dec 2025cs.LGcited by 0

On the map: LLM safety and security → Backdoor and data poisoning attacks

Abstract (arXiv)

This report presents solutions to three machine learning challenges developed as part of the Rayan AI Contest: compositional image retrieval, zero-shot anomaly detection, and backdoored model detection. In compositional image retrieval, we developed a system that processes visual and textual inputs to retrieve relevant images, achieving 95.38% accuracy and ranking first with a clear margin over the second team. For zero-shot anomaly detection, we designed a model that identifies and localizes anomalies in images without prior exposure to abnormal examples, securing second place with a 73.14% score. In the backdoored model detection task, we proposed a method to detect hidden backdoor triggers in neural networks, reaching an accuracy of 78%, which placed our approach in second place. These results demonstrate the effectiveness of our methods in addressing key challenges related to retrieval, anomaly detection, and model security, with implications for real-world applications in industries such as healthcare, manufacturing, and cybersecurity. Code for all solutions is available online (https://github.com/safinal/rayan-ai-contest-solutions).

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Backdoor and data poisoning attacks, one of 12 directions in LLM safety and security. Builds on 9 of its 11 references that are in the map. First posted on arXiv 1 Dec 2025.

Builds on

9 of 11 references.

References · 9

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2023DINOv2: Learning Robust Visual Features without Supervisioncited by 7,974
  3. 2018Representation Learning with Contrastive Predictive Codingcited by 7,777
  4. 2019DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lightercited by 4,307
  5. 2016Identity Mappings in Deep Residual Networkscited by 1,965
  6. 2023AnomalyCLIP: Object-agnostic Prompt Learning for Zero-shot Anomaly Detectioncited by 271
  7. 2022MM-BD: Post-Training Detection of Backdoor Attacks with Arbitrary Backdoor Pattern Types Using a Maximum Margin Statisticcited by 55
  8. 2024MuSc: Zero-Shot Industrial Anomaly Classification and Segmentation with Mutual Scoring of the Unlabeled Imagescited by 52
  9. 2024ViTamin: Designing Scalable Vision Models in the Vision-Language Eracited by 39

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2021EX-RAY: Distinguishing Injected Backdoor from Natural Features in Neural Networks by Examining Differential Feature Symmetry
  2. 2024Model X-ray:Detecting Backdoored Models via Decision Boundary
  3. 2022An anomaly detection approach for backdoored neural networks: face recognition as a case study
  4. 2023SoK: A Systematic Evaluation of Backdoor Trigger Characteristics in Image Classification
  5. 2026RAID: Retrieval-Augmented Anomaly Detection
  6. 2024Advancing Pre-trained Teacher: Towards Robust Feature Discrepancy for Anomaly Detection
  7. 2024Twin Trigger Generative Networks for Backdoor Attacks against Object Detection
  8. 2025Retrieval Augmented Anomaly Detection (RAAD): Nimble Model Adjustment Without Retraining