© 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. 20-TRANSCODER-HP
    7. 141872
    Prev
    Next
    INDEX
    Explanations

    question and answer pairs

    np_max-act · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_20.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.20.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    æĹĥ
    -0.31
     phenomena
    -0.25
     ped
    -0.25
     begging
    -0.24
     serializers
    -0.24
     mir
    -0.24
    饫
    -0.24
    ijing
    -0.24
     pymongo
    -0.24
    以èī²
    -0.24
    POSITIVE LOGITS
     adjust
    0.28
     SECRET
    0.26
    bill
    0.26
    åĩºåıij
    0.26
    -adjust
    0.25
    è°ĥæķ´
    0.25
    hey
    0.25
    éŨ
    0.25
    achel
    0.24
    ä¸įåIJĮ
    0.24
    Activations Density 0.675%

    No Known Activations