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

    Neuronpedia

    Jacobian LensNEW
    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. 38095
    Prev
    Next
    INDEX
    Explanations

    personal descriptions

    np_max-act · gemini-2.0-flash

    engaging and flirtatious language with an emphasis on sexuality.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8

    The neuron responds to the custom persona identifier “NAME_1,” i.e. mentions of the assistant’s own name tag.

    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
    uxtap
    -0.07
    igram
    -0.07
    _print
    -0.06
     Strom
    -0.06
    .prop
    -0.06
    _ur
    -0.06
     розпов
    -0.06
     guests
    -0.06
    isNaN
    -0.06
     Freeman
    -0.06
    POSITIVE LOGITS
     vinegar
    0.07
    алось
    0.06
    ？」
    0.06
    GIT
    0.06
     Medal
    0.06
     cine
    0.06
     dönemde
    0.06
    Yet
    0.06
     rời
    0.06
     parentId
    0.06
    Activations Density 0.026%

    No Known Activations