Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APICircuit TracerNEWSteerSAE EvalsExportsSlackBlogPrivacy & TermsContact
    © Neuronpedia 2025
    Privacy & TermsBlogGitHubSlackTwitterContact
    1. Home
    2. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    3. Llama3.1-8B-IT (Instruct)
    4. Resid Post - 131k
    5. 11-RESID-POST-AA
    6. 80632
    Prev
    Next
    INDEX
    Explanations

    methods

    np_max-act · gemini-2.0-flash

    The neuron activates on non-English/Portuguese tokens—specifically words containing diacritics (e.g. “ção,” “Código”) indicating a request written in Portuguese.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    code related to API endpoints for user registration and protected resource access.

    oai_token-act-pair · gpt-4o-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
    ags
    -0.07
    OutOf
    -0.07
    .getProject
    -0.06
     العق
    -0.06
     fashion
    -0.06
     economics
    -0.06
    .HTML
    -0.06
     GRAT
    -0.06
     Jade
    -0.06
     aided
    -0.06
    POSITIVE LOGITS
    					    
    0.06
     -->↵↵↵
    0.06
    GU
    0.06
     таблет
    0.06
    .MaximizeBox
    0.06
    $$
    0.06
     happen
    0.06
    xFFFFFFFF
    0.06
     zmq
    0.06
    59
    0.06
    Activations Density 0.000%

    No Known Activations