© 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. 27-TRANSCODER-HP
    7. 56871
    Prev
    Next
    INDEX
    Explanations

    statements of fact

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_27.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.27.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    çĿ¿
    -0.27
    æĺ¥é£İ
    -0.26
    strand
    -0.26
    çݺ
    -0.26
    zyst
    -0.26
    è¿ijå¹´
    -0.25
    ,Th
    -0.25
    ÅĽnie
    -0.25
    onym
    -0.25
    èĻİ
    -0.24
    POSITIVE LOGITS
     gatherings
    0.26
    asking
    0.26
     saying
    0.26
     amenities
    0.26
    hear
    0.25
    éĺ³åİ¿
    0.25
    /language
    0.24
    how
    0.24
    aring
    0.24
    çĬĦ
    0.24
    Activations Density 0.006%

    No Known Activations