© 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. 31-TRANSCODER-HP
    7. 18872
    Prev
    Next
    INDEX
    Explanations

    code snippets/terms

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     reckon
    -0.27
    æĪIJé¾Ļ
    -0.26
    åIJįå½ķ
    -0.25
     MMA
    -0.24
    notify
    -0.24
    èĮ«
    -0.24
     Mellon
    -0.24
     slik
    -0.24
    mach
    -0.23
     Sund
    -0.23
    POSITIVE LOGITS
    Term
    0.28
    دت
    0.28
    term
    0.28
    ë°ĺ
    0.27
     Term
    0.27
    åĨ¥
    0.27
    çĶŁåij½
    0.27
    ster
    0.27
    hive
    0.26
    beit
    0.26
    Activations Density 0.842%

    No Known Activations