© 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. 20-TRANSCODER-HP
    7. 130391
    Prev
    Next
    INDEX
    Explanations

    Conjunctions and linking words

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_20.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.20.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     subTitle
    -0.28
    eeper
    -0.27
    peare
    -0.26
    ler
    -0.25
     relative
    -0.24
    ä¹ĭä½ľ
    -0.24
    /********************************************************
    -0.24
    ï¼ģ↵↵↵↵
    -0.24
     late
    -0.24
    mente
    -0.24
    POSITIVE LOGITS
    äºİæĺ¯
    0.31
     thus
    0.29
    æĸ¼æĺ¯
    0.29
    çĦ¶åIJİ
    0.29
    æĤłæĤł
    0.29
    è¿ĻéĩĮæĺ¯
    0.28
    çĦ¶å¾Į
    0.28
    äºİæĺ¯æĪij
    0.28
    çĦ¶èĢĮ
    0.27
     meanwhile
    0.27
    Activations Density 72.389%

    No Known Activations