© 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. 30-TRANSCODER-HP
    7. 96596
    Prev
    Next
    INDEX
    Explanations

    code and computations

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    磨
    -0.31
    ius
    -0.28
    modes
    -0.27
     modes
    -0.26
    ลา
    -0.25
    çĶŁæ´»ä¸Ń
    -0.24
    æıŃ示
    -0.24
    åķ®
    -0.24
    åºķä¸ĭ
    -0.24
     harmon
    -0.24
    POSITIVE LOGITS
    ovable
    0.28
    obar
    0.28
    izedName
    0.27
    åı£æ°´
    0.25
    енÑĮ
    0.25
    år
    0.25
    olang
    0.24
    obot
    0.24
    trimmed
    0.24
     oma
    0.24
    Activations Density 0.003%

    No Known Activations