© 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. 32-TRANSCODER-HP
    7. 43751
    Prev
    Next
    INDEX
    Explanations

    of

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_32.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.32.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     Mood
    -0.28
    éļIJ约
    -0.28
    eco
    -0.27
    iek
    -0.26
    YTE
    -0.26
     frm
    -0.25
     dedication
    -0.25
    .dispatchEvent
    -0.24
     mood
    -0.24
     independently
    -0.23
    POSITIVE LOGITS
    BIG
    0.34
     much
    0.32
    Big
    0.31
     Big
    0.29
     big
    0.29
    å¼Ĥ
    0.29
    è¾ĥ大çļĦ
    0.28
    SION
    0.28
    big
    0.28
    larg
    0.27
    Activations Density 0.002%

    No Known Activations