© 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. 52918
    Prev
    Next
    INDEX
    Explanations

    academic grades

    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
    éĽĨåĽ¢èĤ¡ä»½
    -0.31
    Initializing
    -0.27
    past
    -0.27
    efully
    -0.27
    idis
    -0.26
    à¸Ĥà¸Ļ
    -0.26
    YRO
    -0.26
    mini
    -0.25
    Past
    -0.25
    å¾®åŀĭ
    -0.25
    POSITIVE LOGITS
     grade
    0.33
     top
    0.28
     grad
    0.26
     Grade
    0.25
    åѦç§ij
    0.25
     Degree
    0.24
     score
    0.24
    [Int
    0.24
     './
    0.24
    grade
    0.24
    Activations Density 0.008%

    No Known Activations