© 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. 54034
    Prev
    Next
    INDEX
    Explanations

    become

    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
    berman
    -0.28
     weighting
    -0.26
     пеÑĢвÑĥÑİ
    -0.26
     congr
    -0.25
     bounce
    -0.25
    vere
    -0.25
    ehr
    -0.25
    åĴı
    -0.25
     Say
    -0.24
     Contrib
    -0.24
    POSITIVE LOGITS
    é«ĺåĵģè´¨
    0.28
    ä¸įåIJĮç±»åŀĭ
    0.27
    elia
    0.27
    è¿Ļæĺ¯ä¸Ģ个
    0.27
    ases
    0.26
    Dlg
    0.25
    å°ĨåĨĽ
    0.25
    ants
    0.24
    éĤ¶
    0.24
    letters
    0.24
    Activations Density 0.004%

    No Known Activations