© 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. 108010
    Prev
    Next
    INDEX
    Explanations

    Code/statistics

    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
    emann
    -0.29
    ä¸į失
    -0.26
    PRETTY
    -0.26
    çĶľçĶľ
    -0.26
    PURE
    -0.26
    çĶŁçĶŁ
    -0.26
    æ°ª
    -0.26
     trope
    -0.25
    aman
    -0.25
    失
    -0.24
    POSITIVE LOGITS
    æīįèĥ½
    0.28
    ãģĭãģij
    0.25
     Override
    0.24
     dug
    0.24
    交åıī
    0.24
    工夫
    0.24
    _intersect
    0.23
    Reducer
    0.23
    née
    0.23
    éħįå¥Ĺ
    0.23
    Activations Density 6.577%

    No Known Activations