© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it
    Neuronpedia
    Jacobian LensNEW
    Natural Language
    Autoencoders
    MAY 26    Assistant AxisJAN 26Circuit TracerAPR 26SteerSAE EvalsExportsinterp-engineNEW Community BlogPrivacy & TermsContact
    1. Home
    3. Qwen3-4B
    5. 25-TRANSCODER-HP
    7. 107045
    Prev
    Next
    INDEX
    Explanations

    treat

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_25.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.25.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Negative Logits
     foot
    -0.26
    cre
    -0.26
    aid
    -0.26
     Arrays
    -0.26
     directional
    -0.25
    åĮºéĹ´
    -0.25
    izu
    -0.24
    ="--
    -0.24
    åĨµ
    -0.24
    ä¸įçķĻ
    -0.24
    POSITIVE LOGITS
    æ¡Įä¸Ĭ
    0.25
     viewDidLoad
    0.25
    éĢģä¸Ĭ
    0.25
    PTS
    0.24
    agrid
    0.24
     Viol
    0.24
    .IsAny
    0.24
    datagrid
    0.24
     insurgency
    0.24
    amina
    0.23
    Activations Density 0.022%

    No Known Activations