© 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. 89151
    Prev
    Next
    INDEX
    Explanations
    No Explanations Found
    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
    =count
    -0.29
    =cut
    -0.27
    beg
    -0.26
    ÑĢеб
    -0.26
    ãĤ¸
    -0.24
    ertz
    -0.24
    æŶéĻIJ
    -0.23
    %</
    -0.23
    âĢ¦”↵↵
    -0.23
    erset
    -0.23
    POSITIVE LOGITS
    çĺ¸
    0.29
     specials
    0.25
    骸
    0.25
    jie
    0.25
    ervative
    0.25
    ripper
    0.25
    resume
    0.24
    é¹Ĺ
    0.24
    ixture
    0.24
    ix
    0.24
    Activations Density 10.313%

    No Known Activations

    This feature has no known activations.