© 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. 19302
    Prev
    Next
    INDEX
    Explanations

    blue

    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
     Vac
    -0.29
    vac
    -0.27
    éĹ®
    -0.26
    å·´å·´
    -0.25
    aces
    -0.25
    teness
    -0.25
    TMP
    -0.25
    Thunk
    -0.25
    çģ«
    -0.25
     vac
    -0.25
    POSITIVE LOGITS
    print
    0.38
    prints
    0.34
     print
    0.33
     Prints
    0.33
     prints
    0.32
    PRINT
    0.32
     PRINT
    0.29
    åį°åĪ·
    0.29
    PressEvent
    0.29
    åį°
    0.28
    Activations Density 1.756%

    No Known Activations