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. 81449
    Prev
    Next
    INDEX
    Explanations

    pornography

    np_max-act · gemini-2.0-flash

    The neuron detects fragments of URLs and web addresses (e.g., “http://”, “.com”, “URL” tokens).

    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
     moto
    -0.07
    o
    -0.07
    й
    -0.06
     Sund
    -0.06
     cls
    -0.06
    ks
    -0.06
    ул
    -0.06
    thesized
    -0.06
    �
    -0.06
     perception
    -0.06
    POSITIVE LOGITS
     without
    0.07
    without
    0.06
    μει
    0.06
    ocator
    0.06
    éc
    0.06
    들을
    0.06
     cialis
    0.06
    aimassage
    0.06
     tirelessly
    0.06
    ائي
    0.06
    Activations Density 0.038%

    No Known Activations