© 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. 33-TRANSCODER-HP
    7. 52935
    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_33.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.33.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    Works
    -0.26
    æ¼Ĥ
    -0.25
     responders
    -0.25
     imp
    -0.24
    Cum
    -0.23
    æĬ¬å¤´
    -0.23
    usting
    -0.23
    ä¸ĢåĪĢ
    -0.23
    ¶Ī
    -0.23
    AttributeName
    -0.23
    POSITIVE LOGITS
     keep
    0.41
     Keep
    0.29
     help
    0.29
    keep
    0.29
     KEEP
    0.29
     be
    0.28
     hear
    0.28
    XM
    0.28
    ä¿ĿæĮģ
    0.28
     take
    0.28
    Activations Density 0.013%

    No Known Activations