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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    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. 90711
    Prev
    Next
    INDEX
    Explanations

    history

    np_max-act · gemini-2.0-flash

    The neuron activates on keywords related to historical and cultural educational content—terms like “historical,” “periods,” “events,” “cultures,” and “personalities.”

    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
    icolor
    -0.06
    (wallet
    -0.06
    _offer
    -0.06
     silhouette
    -0.06
    kul
    -0.06
    WAR
    -0.06
    ildi
    -0.06
     Bangkok
    -0.06
    CK
    -0.06
    unset
    -0.06
    POSITIVE LOGITS
    glfw
    0.07
     americ
    0.07
    驶
    0.07
    .entrySet
    0.06
     nour
    0.06
    (Control
    0.06
     grammar
    0.06
    DF
    0.06
    apiKey
    0.06
     più
    0.06
    Activations Density 0.016%

    No Known Activations