© 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
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 12317
    Prev
    Next
    INDEX
    Explanations

    on

    np_max-act · gemini-2.0-flash

    The neuron responds to strong negative or extreme adjectives describing unethical, harmful, or otherwise “over‐the‐top” content.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    andyrdt/saes-llama-3.1-8b-instruct/resid_post_layer_11/trainer_1
    Dataset (Dashboard)
    Various
    Features
    131,072
    Data Type
    float32
    Hook Name
    blocks.11.hook_resid_post
    Architecture
    standard
    Context Size
    1,024
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Marijuana
    -0.07
    че
    -0.06
     LSD
    -0.06
     플레이
    -0.06
    dashboard
    -0.06
    EEK
    -0.06
    ssa
    -0.06
     Vera
    -0.06
    _GPU
    -0.06
    製
    -0.06
    POSITIVE LOGITS
    ellidos
    0.07
     Assert
    0.06
     pledge
    0.06
     Hab
    0.06
     javax
    0.06
    (det
    0.06
     porque
    0.06
     ".";↵
    0.06
     boto
    0.06
    _ix
    0.06
    Activations Density 1.877%

    No Known Activations