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. 81451
    Prev
    Next
    INDEX
    Explanations

    binary search algorithms

    np_max-act · gemini-2.0-flash

    This neuron responds to numeric values (literals) and numeric‐related tokens (e.g. indices, bounds, midpoints) in code.

    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
    iever
    -0.08
    Ingrese
    -0.07
    isis
    -0.07
    iu
    -0.06
     Israelis
    -0.06
    /#{
    -0.06
     AlertDialog
    -0.06
     solidarity
    -0.06
     виде
    -0.06
     unn
    -0.06
    POSITIVE LOGITS
     piping
    0.06
     situated
    0.06
    ATHER
    0.06
     kazanç
    0.06
     Correspond
    0.06
    _feed
    0.06
    _IM
    0.06
    :YES
    0.06
    _version
    0.05
    $action
    0.05
    Activations Density 0.006%

    No Known Activations