© 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. 57048
    Prev
    Next
    INDEX
    Explanations

    assistant

    np_max-act · gemini-2.0-flash

    The neuron is detecting the Spanish phrase “asistente médico” (i.e. when the text specifies “you are a medical assistant”).

    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
     Matthias
    -0.07
    -bootstrap
    -0.07
    _successful
    -0.07
    ANTI
    -0.06
    -notch
    -0.06
     taraf
    -0.06
     mã
    -0.06
    _cpu
    -0.06
    keiten
    -0.06
    刻
    -0.06
    POSITIVE LOGITS
    hot
    0.06
    *y
    0.06
     fecha
    0.06
     '[
    0.06
    _mapping
    0.06
    pray
    0.06
    bam
    0.06
     stark
    0.06
     Bryce
    0.06
     Absolutely
    0.06
    Activations Density 0.013%

    No Known Activations