© 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. 33-TRANSCODER-HP
    7. 53820
    Prev
    Next
    INDEX
    Explanations

    Rescues and disasters

    np_max-act-logits · gemini-2.0-flash
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    mwhanna/qwen3-4b-transcoders/layer_33.safetensors
    Prompts (Dashboard)
    16,384 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    163,840
    Data Type
    float32
    Hook Name
    blocks.33.mlp.hook_in
    Architecture
    transcoder
    Context Size
    8,192
    Dataset
    monology/pile-uncopyrighted
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Negative Logits
    eteor
    -0.28
    bred
    -0.27
    heten
    -0.24
     fed
    -0.24
    -fed
    -0.24
     senior
    -0.23
     congrat
    -0.23
     bols
    -0.23
     me
    -0.23
    chaft
    -0.23
    POSITIVE LOGITS
    spots
    0.28
    æīĵæĸŃ
    0.28
    ubic
    0.27
    æīĵä»Ĺ
    0.27
    åģ¥
    0.26
    lä
    0.26
    å´©æºĥ
    0.26
    é©¾æł¡
    0.25
    åı¸æľº
    0.25
    éĻįä½İæĪIJæľ¬
    0.25
    Activations Density 0.021%

    No Known Activations