© 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. 22-TRANSCODER-HP
    7. 98507
    Prev
    Next
    INDEX
    Explanations

    we

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_22.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.22.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    èı
    -0.27
    æĺ¯ä½ł
    -0.26
    /train
    -0.26
    åĪĨæķ£
    -0.24
    taken
    -0.24
    adopt
    -0.23
    uffle
    -0.23
    èĩªç§ģ
    -0.23
    resher
    -0.23
    epar
    -0.23
    POSITIVE LOGITS
    ä¹Łåıªèĥ½
    0.32
    owie
    0.30
    >Type
    0.25
    mlink
    0.25
    vos
    0.24
    è¿ĩåIJİ
    0.24
     wonder
    0.24
    oun
    0.24
    åıªèĥ½
    0.24
    UIL
    0.23
    Activations Density 0.003%

    No Known Activations