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. 84748
    Prev
    Next
    INDEX
    Explanations

    code/system commands

    np_max-act · gemini-2.0-flash

    The neuron responds to special control and formatting tokens (e.g. “|”, “im”, “_start”, “_end,” and other model-internal markers) used to delimit or annotate sections of the text.

    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
     versión
    -0.06
     exam
    -0.06
    従
    -0.06
     환산
    -0.06
     courseId
    -0.06
     Ά
    -0.06
    ダー
    -0.06
     جستارهای
    -0.06
    .radians
    -0.06
    Whats
    -0.06
    POSITIVE LOGITS
    QueryString
    0.06
     Sylv
    0.06
    eggies
    0.06
    ymce
    0.06
    /");↵
    0.06
    つけ
    0.06
     ipv
    0.06
     enctype
    0.06
    orest
    0.06
    CallBack
    0.06
    Activations Density 0.010%

    No Known Activations