© 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. 108795
    Prev
    Next
    INDEX
    Explanations

    current

    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
    çİī
    -0.27
     glo
    -0.24
     childhood
    -0.24
    ember
    -0.24
    <TSource
    -0.24
    ERY
    -0.24
    è¯Ĺ
    -0.23
    äºij端
    -0.23
    æŀļ
    -0.23
    è¦Ĩ
    -0.23
    POSITIVE LOGITS
     affairs
    0.29
     flowing
    0.28
    astle
    0.28
     Affairs
    0.27
     densities
    0.27
    aned
    0.26
     Westbrook
    0.26
     assignable
    0.26
    strup
    0.25
    sport
    0.25
    Activations Density 1.337%

    No Known Activations