Llama 3.2 1B Instruct abliterated

Llama

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

Released October 2024131K context≈ 98,304 words1B params

A stripped-down, uncensored variant of Meta's Llama 3.2 1B that has had its refusal behaviors removed through abliteration — a technique that suppresses the model's trained tendency to decline requests. It operates without the safety guardrails present in the base instruct model, meaning it will attempt to respond to a much wider range of prompts. Being a 1B parameter model, its raw reasoning capacity is limited compared to larger models.

Capabilities

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

Multilingual

Moderate

Tool Use

Llama 3.2 1B Instruct abliterated

Llama

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

Released October 2024131K context≈ 98,304 words1B params

A stripped-down, uncensored variant of Meta's Llama 3.2 1B that has had its refusal behaviors removed through abliteration — a technique that suppresses the model's trained tendency to decline requests. It operates without the safety guardrails present in the base instruct model, meaning it will attempt to respond to a much wider range of prompts. Being a 1B parameter model, its raw reasoning capacity is limited compared to larger models.

Capabilities

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

Multilingual

Moderate

Tool Use

Glossary

AbliterationA technique that removes or disables a model's built-in safety refusal mechanisms, allowing it to respond to a wider range of requests.GuardrailsSafety mechanisms built into a model to refuse harmful requests or prevent it from generating unsafe content.Parameter ModelA neural network described by the number of learnable weights it contains; more parameters generally mean greater capacity to learn complex patterns, but also require more computational resources.ReasoningThe model's ability to work through multi-step logical problems and provide justified answers rather than just pattern-matching.Reasoning CapacityThe model's ability to perform complex logical thinking and problem-solving tasks beyond simple pattern matching.Safety GuardrailsBuilt-in restrictions or filters that prevent a model from generating harmful, illegal, or unethical content.UncensoredA model without built-in safety filters or content restrictions, allowing it to generate responses on any topic without refusal.

Capabilities

Capabilities

Use Case Fit

Glossary