© 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. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 43026
    Prev
    Next
    INDEX
    Explanations

    keep

    np_max-act · gemini-2.0-flash

    The neuron fires on words and phrases used in the assistant’s safety-and-ethics guidance (e.g. “safe,” “harmful,” “appropriate,” “ethical,” “responsible”).

    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
     panoramic
    -0.07
     onun
    -0.07
    ?")↵
    -0.07
     week
    -0.06
    HttpServletRequest
    -0.06
    )?
    -0.06
    ‘
    -0.06
    _sell
    -0.06
     Số
    -0.06
     quảng
    -0.06
    POSITIVE LOGITS
    erli
    0.07
     seab
    0.06
    華
    0.06
    ="[
    0.06
    cls
    0.06
    .assertIs
    0.06
    클
    0.06
    (newState
    0.06
     CoreData
    0.06
    드
    0.05
    Activations Density 0.424%

    No Known Activations