© 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. 18-TRANSCODER-HP
    7. 88814
    Prev
    Next
    INDEX
    Explanations

    Code or documentation

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_18.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.18.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     lifts
    -0.27
    å°Ķæĸ¯
    -0.25
     flies
    -0.24
     introducing
    -0.24
     Flying
    -0.24
     tensions
    -0.24
    é£ŀ
    -0.23
    levator
    -0.23
     Splash
    -0.23
     superf
    -0.23
    POSITIVE LOGITS
    ersen
    0.27
    spo
    0.26
    cken
    0.25
     .$
    0.25
    满
    0.25
    GRAM
    0.25
    едÑĮ
    0.24
    rän
    0.24
     mol
    0.24
    oct
    0.24
    Activations Density 1.101%

    No Known Activations