© 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. 63784
    Prev
    Next
    INDEX
    Explanations

    random, disparate texts

    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
    ecc
    -0.28
    ä¸įåĪĨ
    -0.26
    åºĶåıĬæŶ
    -0.25
    åĻĻ
    -0.24
    ä¸įæŃ»
    -0.24
    å¼ĤæĢ§
    -0.24
    åĪ©çī©
    -0.23
    avras
    -0.23
    .should
    -0.23
     —↵↵
    -0.23
    POSITIVE LOGITS
    èģĶåIJĪ
    0.36
    union
    0.30
    åĪĨæķ£
    0.30
    hit
    0.27
     hits
    0.27
    çļĦä¸ĢäºĽ
    0.26
     union
    0.25
    poly
    0.25
     hit
    0.25
    ploy
    0.25
    Activations Density 3.222%

    No Known Activations