© 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. 54001
    Prev
    Next
    INDEX
    Explanations

    code and citations

    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
    า
    -0.33
     PAN
    -0.29
     incon
    -0.28
    ieten
    -0.26
     ton
    -0.26
    ier
    -0.26
    a
    -0.26
     required
    -0.26
     appropriate
    -0.25
     intr
    -0.25
    POSITIVE LOGITS
    rets
    0.30
    å¼¹
    0.28
    AsString
    0.27
    åºł
    0.27
    表éĿ¢ä¸Ĭ
    0.26
    =__
    0.26
    æĬ¢åħĪ
    0.25
     Compilation
    0.24
    NL
    0.24
    åIJįçīĮ
    0.24
    Activations Density 0.003%

    No Known Activations