© 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. 31-TRANSCODER-HP
    7. 18899
    Prev
    Next
    INDEX
    Explanations

    Locations

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    arius
    -0.29
    /IP
    -0.27
    bp
    -0.27
    ÑģпекÑĤ
    -0.26
     Tristan
    -0.26
     document
    -0.25
    æĸĩæ¡£
    -0.25
     manh
    -0.25
    /document
    -0.25
    åıĹ害èĢħ
    -0.25
    POSITIVE LOGITS
    ény
    0.26
    çŃı
    0.26
     Sandbox
    0.25
    MDB
    0.25
    _('
    0.24
    注æĺİåĩºå¤Ħ
    0.24
    ahoma
    0.24
    ooth
    0.24
     Merrill
    0.24
    atically
    0.24
    Activations Density 0.005%

    No Known Activations