© 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. 107091
    Prev
    Next
    INDEX
    Explanations

    Allow

    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
    çĶ¸
    -0.30
    ç½®ä¸ļ
    -0.26
     preset
    -0.25
    è´«
    -0.25
     Hardcore
    -0.25
     initialised
    -0.24
    åĻľ
    -0.24
    åı¯ä»¥ä½¿
    -0.24
     Fresh
    -0.24
     Pun
    -0.24
    POSITIVE LOGITS
    losure
    0.26
    æī«ä¸Ģæī«
    0.24
    uffy
    0.24
    çķ¥æľī
    0.24
    ynchronously
    0.24
     &:
    0.23
    è¿Ĥ
    0.23
    leur
    0.23
    .Try
    0.23
     maç
    0.23
    Activations Density 9.094%

    No Known Activations