HarmBench Llama 2 13b cls

Llama 2

Open WeightModel weights are publicly available — can be downloaded and self-hosted

Released February 20242K context≈ 1,536 words13b params

A specialized classifier built to detect harmful content rather than generate it. This model operates as a judge, evaluating whether text crosses safety boundaries — it reads inputs and labels them rather than conversing. Its narrow focus makes it reliable within its lane but unsuitable for general-purpose tasks.

Capabilities

Capability scores are AI-generated based on model documentation, benchmarks, and technical specifications. Learn more

Multilingual

Moderate

Long Context

HarmBench Llama 2 13b cls

Llama 2

Open WeightModel weights are publicly available — can be downloaded and self-hosted

Released February 20242K context≈ 1,536 words13b params

A specialized classifier built to detect harmful content rather than generate it. This model operates as a judge, evaluating whether text crosses safety boundaries — it reads inputs and labels them rather than conversing. Its narrow focus makes it reliable within its lane but unsuitable for general-purpose tasks.

Capabilities

Capability scores are AI-generated based on model documentation, benchmarks, and technical specifications. Learn more

Multilingual

Moderate

Long Context

Glossary

ClassifierA machine learning model trained to assign input data into predefined categories or labels.General-PurposeDesigned to handle a wide variety of different tasks rather than being specialized for one specific domain.

HarmBench Llama 2 13b cls

Capabilities

HarmBench Llama 2 13b cls

Capabilities

Use Case Fit

Similar Models

Glossary