© 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. 19327
    Prev
    Next
    INDEX
    Explanations

    than

    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
    erton
    -0.28
    arat
    -0.28
     Portions
    -0.26
    è¿ĻäºĽä¸ľè¥¿
    -0.26
    çļĦæ¯Ķä¾ĭ
    -0.26
    åĿļ
    -0.25
    andle
    -0.25
    ä¸ĩæĪ·
    -0.24
    bat
    -0.24
    象
    -0.24
    POSITIVE LOGITS
     that
    0.31
     being
    0.30
     just
    0.30
    JUST
    0.29
     we
    0.29
     regular
    0.29
    ä»Ĭ天æĪij们
    0.28
    just
    0.28
     twice
    0.28
     lo
    0.28
    Activations Density 0.140%

    No Known Activations