© 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. 54407
    Prev
    Next
    INDEX
    Explanations

    Non-English text

    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
    que
    -0.30
    ESS
    -0.27
     quant
    -0.26
    èħ®
    -0.25
    EST
    -0.25
    ToLeft
    -0.25
     Cou
    -0.25
     cou
    -0.25
     bell
    -0.25
    Cached
    -0.24
    POSITIVE LOGITS
    hus
    0.26
    tu
    0.24
    å¹´ä¸ĭåįĬå¹´
    0.24
    ãģ¡
    0.24
    stub
    0.24
    ture
    0.24
    Fl
    0.24
    æĥħåĨµè¿Ľè¡Į
    0.24
     Há
    0.23
    lug
    0.23
    Activations Density 0.017%

    No Known Activations