© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. OpenMOSS · Llama Scope: SAEs for Llama-3.1-8B
    5. Llama3.1-8B (Base)
    7. Residual Stream
    9. 17-LLAMASCOPE-RES-131K
    11. 82431
    Prev
    Next
    INDEX
    Explanations

    words or phrases associated with severe negative impacts or consequences

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    cerebras/SlimPajama-627B
    No Configuration Found
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    osen
    -0.07
    sez
    -0.07
    isle
    -0.07
    бÑĥ
    -0.07
    allet
    -0.07
    çok
    -0.07
    ettle
    -0.06
     kans
    -0.06
    elier
    -0.06
    anna
    -0.06
    POSITIVE LOGITS
     consequences
    0.08
    /dev
    0.08
    antly
    0.07
    COPE
    0.07
     loss
    0.07
    cope
    0.07
    lest
    0.07
    anches
    0.07
    LY
    0.07
    оба
    0.07
    Activations Density 0.008%

    No Known Activations