© 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. 18952
    Prev
    Next
    INDEX
    Explanations

    on

    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
    çīĩ
    -0.28
     illegally
    -0.28
    .at
    -0.28
     at
    -0.28
    istically
    -0.26
    doch
    -0.26
    å°ĨæĿ¥
    -0.25
    еÑĩно
    -0.25
     mxArray
    -0.25
     violently
    -0.24
    POSITIVE LOGITS
    éĤ£ä¹Īå¤ļ
    0.33
    è¿Ļä¹Īå¤ļ
    0.33
     so
    0.33
    multiple
    0.32
    好å¤ļ
    0.32
     multiple
    0.31
     lower
    0.30
    Multiple
    0.29
    å¤ļä½į
    0.29
     So
    0.28
    Activations Density 0.021%

    No Known Activations