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

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & 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. 31
    Prev
    Next
    INDEX
    Explanations

    comma

    np_max-act · gemini-2.0-flash

    the neuron responds to explicit sexual content, especially direct references to genitals and erotic acts.

    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
    méně
    -0.07
     freund
    -0.06
    ์เพ
    -0.06
    _recipe
    -0.06
    -fly
    -0.06
     عملی
    -0.06
    .UltraWin
    -0.06
    olesale
    -0.06
    φερ
    -0.06
    ktop
    -0.06
    POSITIVE LOGITS
     Lak
    0.07
     Rol
    0.06
    (..
    0.06
    (typeof
    0.06
     (?)
    0.06
     (_,
    0.06
     typeof
    0.06
    Vm
    0.06
    :m
    0.06
     أر
    0.06
    Activations Density 0.088%

    No Known Activations