© 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. 141396
    Prev
    Next
    INDEX
    Explanations

    of

    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
    åĩı
    -0.31
    mana
    -0.26
    leftrightarrow
    -0.24
     PolÃŃtica
    -0.24
     przez
    -0.24
    amina
    -0.23
    _future
    -0.23
    æ¸Ĺ
    -0.23
    att
    -0.23
     displ
    -0.23
    POSITIVE LOGITS
    evin
    0.30
    头éĥ¨
    0.29
    éĥ½è¢«
    0.28
    ramid
    0.28
    emens
    0.27
    èį¯å¸Ī
    0.27
    )row
    0.25
    ?id
    0.25
    åįĩèµ·
    0.25
    çľĭäºĨä¸Ģ
    0.24
    Activations Density 13.159%

    No Known Activations