© 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. 107034
    Prev
    Next
    INDEX
    Explanations

    rather

    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
    aneously
    -0.31
    adero
    -0.30
    endor
    -0.27
    ially
    -0.26
    USES
    -0.26
    ä¾®
    -0.25
    Raster
    -0.25
    arer
    -0.25
    statusCode
    -0.24
    rary
    -0.24
    POSITIVE LOGITS
     having
    0.43
    having
    0.35
     Having
    0.35
     being
    0.32
     needing
    0.31
    Having
    0.30
     realizing
    0.30
     doing
    0.30
     offering
    0.29
     paying
    0.29
    Activations Density 1.076%

    No Known Activations