© 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. 141902
    Prev
    Next
    INDEX
    Explanations

    Raw text extracts

    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
    è°ķ
    -0.28
    éné
    -0.25
    è¯Ŀ说
    -0.25
    çĶľç¾İ
    -0.24
    åij½åIJį
    -0.24
    åĽ½å®¶éĺŁ
    -0.24
    Narr
    -0.24
    åĽ½åĬ¡
    -0.23
    æ»ĭåij³
    -0.23
    çĿij
    -0.23
    POSITIVE LOGITS
    -slot
    0.26
    第äºĮå±Ĭ
    0.24
    alle
    0.24
    éĽĨåĽ¢
    0.24
    ally
    0.24
    èij£äºĭä¼ļ
    0.23
    ãģ¦ãĤĭ
    0.23
    å¾Īå·®
    0.23
    slot
    0.23
    ibri
    0.23
    Activations Density 4.661%

    No Known Activations