© 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. 18146
    Prev
    Next
    INDEX
    Explanations

    in

    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
     lut
    -0.31
    è´¯éĢļ
    -0.29
     cry
    -0.26
    <textarea
    -0.25
     restr
    -0.25
    æĥ¯
    -0.25
     pet
    -0.24
    çĸı
    -0.24
    é¦ĸ
    -0.24
     hab
    -0.24
    POSITIVE LOGITS
     boxes
    0.45
     Box
    0.44
    Box
    0.43
     box
    0.42
     Boxes
    0.42
    box
    0.40
    _box
    0.40
    -box
    0.40
    BOX
    0.38
     BOX
    0.37
    Activations Density 1.255%

    No Known Activations