AnoLLM: large language models for tabular anomaly detection

AnoLLM uses large language models to detect anomalies in mixed-type tabular data by converting rows to a standardized text format and scoring records with negative log likelihood.

AnoLLM is a framework that leverages large language models for unsupervised tabular anomaly detection. The method converts tabular records into a standardized textual serialization, then adapts a pre-trained large language model with that serialized data. AnoLLM assigns anomaly scores based on the negative log likelihood produced by the adapted model. The authors position this approach as an alternative to traditional tabular methods that often require extensive feature engineering and can strip or lose textual information during preprocessing.

The framework emphasizes preserving data integrity and streamlining preprocessing for mixed-type datasets. By serializing rows into text, AnoLLM can naturally incorporate textual features alongside numerical and categorical fields without separate feature transformations. The approach adapts a pre-trained model to the serialized tabular format rather than building specialized handcrafted encoders for each column type. That design choice is intended to reduce the labor and complexity of feature engineering while retaining textual signals that can be informative for anomaly detection.

The authors report empirical results showing AnoLLM delivers the best performance on six benchmark datasets that contain mixed feature types. They also evaluate the method across 30 datasets from the ODDS library, which the article describes as predominantly numerical, and find that AnoLLM performs on par with top performing baselines on those datasets. Taken together, the results indicate the approach is especially effective where textual and mixed-type data are present and remains competitive on primarily numerical collections, offering a unified, model-driven alternative to conventional tabular anomaly detection pipelines.

55

Impact Score

LLM-PIEval: a benchmark for indirect prompt injection attacks in large language models

Large language models have increased interest in Artificial Intelligence and their integration with external tools introduces risks such as direct and indirect prompt injection. LLM-PIEval provides a framework and test set to measure indirect prompt injection risk and the authors release API specifications and prompts to support wider assessment.

NVIDIA may stop bundling memory with gpu kits amid gddr shortage

NVIDIA is reportedly considering supplying only bare silicon to its aic partners rather than the usual gpu and memory kit as gddr shortages constrain fulfillment. The move follows wider industry pressure from soaring dram prices and an impending price increase from AMD of about 10% across its gpu lineup.

SK Hynix to showcase 48 Gb/s 24 Gb GDDR7 for Artificial Intelligence inference

SK Hynix will present a 24 Gb GDDR7 chip rated for 48 Gb/s at ISSCC 2026, claiming a symmetric dual-channel design and updated internal interfaces that push past the expected 32 to 37 Gb/s. The paper positions the device for mid-range Artificial Intelligence inference and SK Hynix will also show LPDDR6 running at 14.4 Gb/s.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.