© 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. 30-TRANSCODER-HP
    7. 96061
    Prev
    Next
    INDEX
    Explanations

    Random text snippets

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_30.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.30.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     unit
    -0.30
     Grid
    -0.29
     grid
    -0.29
    æĭ¨
    -0.28
    ponder
    -0.26
     dist
    -0.25
    é©»
    -0.25
    è·¯åı£
    -0.25
    Grid
    -0.25
    ayed
    -0.24
    POSITIVE LOGITS
    als
    0.29
    _pe
    0.27
    Pe
    0.27
    pe
    0.26
    wall
    0.26
    olas
    0.26
    iets
    0.26
    pee
    0.25
    ports
    0.24
    opia
    0.24
    Activations Density 0.101%

    No Known Activations