© 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. 95590
    Prev
    Next
    INDEX
    Explanations

    removing layers/parts

    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
    åĪĴ
    -0.27
    åĪº
    -0.27
    åĬĥ
    -0.26
     chuck
    -0.25
    agi
    -0.25
    æĭ³å¤´
    -0.24
    achel
    -0.24
    夹
    -0.24
    =~
    -0.24
    oker
    -0.23
    POSITIVE LOGITS
    顺
    0.29
     backward
    0.28
     inside
    0.27
    ä»İåīį
    0.27
     back
    0.27
    çĶļ
    0.26
    [from
    0.25
    ä»İ
    0.25
    toLocale
    0.25
    ston
    0.25
    Activations Density 0.104%

    No Known Activations