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

    Neuronpedia

    APIAssistant AxisNEWCircuit 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. 84647
    Prev
    Next
    INDEX
    Explanations

    male genitalia

    np_max-act · gemini-2.0-flash

    The neuron responds to explicit sexual terms, especially slang or anatomical words for genitalia.

    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
    _BT
    -0.08
     kön
    -0.07
    yl
    -0.06
     Shader
    -0.06
    어
    -0.06
    activity
    -0.06
     rost
    -0.06
     impossible
    -0.06
    ador
    -0.06
    angel
    -0.06
    POSITIVE LOGITS
     RECE
    0.08
    正常
    0.07
     nightmare
    0.06
     tariffs
    0.06
    ifact
    0.06
     Countdown
    0.06
     supply
    0.06
    querySelector
    0.06
    _variables
    0.06
    hesive
    0.06
    Activations Density 0.042%

    No Known Activations