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

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 27-TRANSCODER-HP
    7. 57003
    Prev
    Next
    INDEX
    Explanations

    say the ending -ing,-ive,-ly,-ities,-ier

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_27.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.27.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    обÑĢа
    -0.28
    stown
    -0.26
    ums
    -0.26
    ovie
    -0.25
    beer
    -0.24
    /GPL
    -0.24
    orative
    -0.24
    åĪĨæĺİ
    -0.23
     accru
    -0.23
    iration
    -0.23
    POSITIVE LOGITS
    åĢį
    0.30
    pec
    0.28
    (es
    0.28
    ä¸Ģ大
    0.27
    nosti
    0.26
    -addons
    0.26
    erved
    0.26
    ensitive
    0.25
    ical
    0.25
    ifying
    0.24
    Activations Density 0.014%

    No Known Activations