© 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. 19259
    Prev
    Next
    INDEX
    Explanations

    web

    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
    æĻĶ
    -0.26
     Central
    -0.26
    è¾¾æłĩ
    -0.24
    ä½İ端
    -0.24
    GRADE
    -0.24
     Epid
    -0.24
    íĭ´
    -0.24
    ä¹°åįķ
    -0.24
    çķħéĢļ
    -0.23
     Beste
    -0.23
    POSITIVE LOGITS
     dig
    0.27
    æķ°åŃĹ
    0.26
    illery
    0.25
    ymmetric
    0.25
     digit
    0.25
    -pres
    0.24
    seg
    0.24
    folio
    0.24
     Regents
    0.24
     digital
    0.24
    Activations Density 0.118%

    No Known Activations