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. 86009
    Prev
    Next
    INDEX
    Explanations

    Places and names

    np_max-act · gemini-2.0-flash

    The neuron fires on subword sequences that form capitalized proper names (especially place names or titled entities).

    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
     福
    -0.07
    ponsible
    -0.06
     strap
    -0.06
     string
    -0.06
     programas
    -0.06
    alties
    -0.06
    teří
    -0.06
    .ent
    -0.06
     Horton
    -0.06
     underage
    -0.06
    POSITIVE LOGITS
    صل
    0.06
    aterial
    0.06
    inished
    0.06
     відповід
    0.06
    ット
    0.06
    少年
    0.06
    [test
    0.06
    .Metro
    0.06
    -clean
    0.06
     Tomb
    0.06
    Activations Density 0.081%

    No Known Activations