© 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. 88621
    Prev
    Next
    INDEX
    Explanations

    URLs and HTML

    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
     eens
    -0.27
    åı°ä¸Ĭ
    -0.25
    ighthouse
    -0.24
    amer
    -0.24
     Fold
    -0.23
    orum
    -0.23
    Fold
    -0.23
    /value
    -0.23
     fold
    -0.23
    aise
    -0.23
    POSITIVE LOGITS
     vowels
    0.29
    iclass
    0.27
    rı
    0.26
    adients
    0.25
    åŃº
    0.25
    vio
    0.25
    tet
    0.24
    ä½ľçĶ¨
    0.24
    ket
    0.24
    lene
    0.24
    Activations Density 3.719%

    No Known Activations