© 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. 142203
    Prev
    Next
    INDEX
    Explanations

    Excerpts from written sources

    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
    INIT
    -0.28
    æīįç®Ĺ
    -0.26
    anja
    -0.25
    绿水
    -0.25
    |i
    -0.24
    OOM
    -0.24
    ORIZATION
    -0.24
    å°ıäºĭ
    -0.23
    PIN
    -0.23
    (candidate
    -0.23
    POSITIVE LOGITS
     Dur
    0.27
    宫
    0.27
    Dur
    0.26
    åķĨ
    0.25
    ç§Ĵ
    0.25
    第ä¸ī
    0.25
     inconsist
    0.25
    ple
    0.24
     consistent
    0.24
    第äºĮ
    0.24
    Activations Density 4.066%

    No Known Activations