© 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. 108931
    Prev
    Next
    INDEX
    Explanations

    diverse texts

    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
    éĩĮç¨ĭ
    -0.29
    ammer
    -0.29
    æĦŁæ¿Ģ
    -0.28
    hole
    -0.27
    yo
    -0.26
    intel
    -0.25
    mile
    -0.24
     instrument
    -0.24
    å´Ń
    -0.24
    ynthesis
    -0.24
    POSITIVE LOGITS
    Virgin
    0.33
     Virgin
    0.33
     Tay
    0.27
    å¹³çŃī
    0.27
     virgin
    0.27
    typ
    0.26
    ej
    0.26
    lor
    0.26
     Diy
    0.25
    clair
    0.25
    Activations Density 0.896%

    No Known Activations