© 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. 73837
    Prev
    Next
    INDEX
    Explanations

    code settings

    np_max-act · gemini-2.0-flash

    The neuron activates on CamelCase or mixed-case identifier subwords that start with an uppercase letter (e.g. “Keep,” “Flag,” “No,” “Special,” “Unity,” “Virtual,” etc.).

    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
     impressions
    -0.07
    .adjust
    -0.06
    _callbacks
    -0.06
     About
    -0.06
     Sunder
    -0.06
    ,
    -0.06
    vin
    -0.06
    auge
    -0.06
     eighth
    -0.06
     до
    -0.06
    POSITIVE LOGITS
     комп
    0.07
     Assert
    0.07
    Š
    0.06
     insiders
    0.06
     मन
    0.06
    İK
    0.06
    _ge
    0.06
     retorno
    0.06
     //@
    0.06
    initialized
    0.06
    Activations Density 0.045%

    No Known Activations