© 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. 31-TRANSCODER-HP
    7. 19545
    Prev
    Next
    INDEX
    Explanations

    something getting better or worse

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_31.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.31.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
     contracts
    -0.27
    hya
    -0.27
     not
    -0.26
    anian
    -0.26
     wheels
    -0.26
    psc
    -0.26
    agna
    -0.25
    hatt
    -0.24
     pik
    -0.24
     enhancements
    -0.23
    POSITIVE LOGITS
    little
    0.31
     little
    0.30
    左侧
    0.29
    lost
    0.28
    æ¶ķ
    0.28
    elder
    0.28
    æ¡´
    0.27
    æĪª
    0.27
     left
    0.26
    outed
    0.26
    Activations Density 0.886%

    No Known Activations