© 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. 95317
    Prev
    Next
    INDEX
    Explanations

    jewelry

    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
    è¯Ŀ
    -0.28
    å¼Ģåıij建设
    -0.27
    çŃĨ
    -0.26
    å§Ķç»Ħç»ĩ
    -0.25
    ister
    -0.25
    éĺŁä¼į建设
    -0.24
    achinery
    -0.24
    privileged
    -0.24
    çļĦæĸĩåŃĹ
    -0.24
     princ
    -0.23
    POSITIVE LOGITS
    omet
    0.31
     bench
    0.30
    stone
    0.30
    bench
    0.28
    ORE
    0.27
    wear
    0.26
     maker
    0.26
    -design
    0.26
     designs
    0.26
     Bench
    0.26
    Activations Density 0.887%

    No Known Activations