© 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. 19354
    Prev
    Next
    INDEX
    Explanations

    carcinoma

    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
    æ³Ħæ¼ı
    -0.29
    çļĦ年轻人
    -0.26
    ppe
    -0.25
    кан
    -0.25
    èµ°ä¸ĭåİ»
    -0.25
    è¡Ģèħ¥
    -0.25
     leakage
    -0.24
     twig
    -0.24
     ful
    -0.24
    PathParam
    -0.24
    POSITIVE LOGITS
     Foundation
    0.31
    çļĦåŁºç¡Ģ
    0.28
    åŁº
    0.27
     basal
    0.27
    맹
    0.27
     Bas
    0.27
    Foundation
    0.26
    Bas
    0.26
     Foundations
    0.25
    /Base
    0.25
    Activations Density 0.998%

    No Known Activations