© 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. 25-TRANSCODER-HP
    7. 108679
    Prev
    Next
    INDEX
    Explanations

    prepositions

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_25.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.25.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     stranded
    -0.26
    æĽĿ
    -0.24
    -su
    -0.24
    plr
    -0.24
    /print
    -0.24
    ãĥ«ãĥī
    -0.23
     exponentially
    -0.23
    群éĩĮ
    -0.23
     Printable
    -0.23
    uary
    -0.23
    POSITIVE LOGITS
    ÑĢаг
    0.31
    oka
    0.30
    èĪĴéĢĤ
    0.27
    oma
    0.27
    ocoa
    0.26
    oke
    0.26
    ispens
    0.25
    oha
    0.25
    _OP
    0.24
    iman
    0.24
    Activations Density 0.938%

    No Known Activations