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. 40644
    Prev
    Next
    INDEX
    Explanations

    Informal, conversational text

    np_max-act · gemini-2.0-flash

    This neuron detects dramatic or fantasy‐style “X of Y” name phrases (e.g. “moons of …,” “goddess of …”).

    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
    "If
    -0.07
    &apos
    -0.06
    Pokud
    -0.06
    äter
    -0.06
     overwhelmed
    -0.06
     acquaint
    -0.06
    Từ
    -0.06
     viewed
    -0.06
     طرح
    -0.06
    Trying
    -0.06
    POSITIVE LOGITS
     Λ
    0.08
    (forms
    0.07
    ниць
    0.06
     biçim
    0.06
     alan
    0.06
    .console
    0.06
     Failure
    0.06
    alue
    0.06
     Oasis
    0.06
     Grinding
    0.06
    Activations Density 0.022%

    No Known Activations