© 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. 56953
    Prev
    Next
    INDEX
    Explanations

    to

    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
    åĪ»
    -0.32
     Mixer
    -0.32
    辩
    -0.29
    abs
    -0.28
    .Toggle
    -0.27
    éĩįå·¥
    -0.26
    üh
    -0.26
    enez
    -0.26
    ulu
    -0.26
    æ¶ī
    -0.25
    POSITIVE LOGITS
     softer
    0.27
    ropol
    0.26
    /doc
    0.26
    à¸ĩาà¸Ļ
    0.25
     staying
    0.25
     keeping
    0.25
    ĶåĽŀ
    0.25
     focus
    0.25
    VILLE
    0.24
     dealing
    0.24
    Activations Density 0.004%

    No Known Activations