© 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. 89398
    Prev
    Next
    INDEX
    Explanations

    to

    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
     Leaves
    -0.28
    è¾¹
    -0.26
    æĥ¹
    -0.26
    ç»Ħè£ħ
    -0.25
     Ori
    -0.25
    iÄĩ
    -0.25
    ãģĻãģIJ
    -0.25
    areth
    -0.24
    igion
    -0.24
    ieved
    -0.24
    POSITIVE LOGITS
    /html
    0.27
    unker
    0.27
    æľīä¸Ģä½į
    0.27
    tem
    0.26
    è¾¾æłĩ
    0.26
     core
    0.26
     modular
    0.26
    éné
    0.25
    æĹºçĽĽ
    0.25
    cing
    0.25
    Activations Density 13.405%

    No Known Activations