© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Andy Arditi · Finding Misaligned Persona Features in Open-Weight Models
    5. Llama3.1-8B-IT (Instruct)
    7. Resid Post - 131k
    9. 11-RESID-POST-AA
    11. 109664
    Prev
    Next
    INDEX
    Explanations

    Wild

    np_max-act · gemini-2.0-flash

    This neuron detects occurrences of the standalone word “Wild,” especially when it appears as the first token in names or titles.

    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
     格
    -0.08
    ソン
    -0.08
    $con
    -0.07
     Morton
    -0.07
     Mehmet
    -0.07
     Düş
    -0.07
     договор
    -0.07
     uncomfort
    -0.07
    percentage
    -0.07
     decreased
    -0.07
    POSITIVE LOGITS
     Wild
    0.16
     wild
    0.14
    Wild
    0.12
    wild
    0.10
    野
    0.09
     Wilde
    0.09
     Wildcats
    0.09
     Wilderness
    0.08
     wildfire
    0.08
    ILD
    0.08
    Activations Density 0.010%

    No Known Activations