© 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. 63644
    Prev
    Next
    INDEX
    Explanations

    Code, numbers, and foreign languages

    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
    Outside
    -0.30
     Outside
    -0.26
    以èī²
    -0.26
    æijĦå½±ä½ľåĵģ
    -0.25
    å³Ĵ
    -0.25
    outside
    -0.25
     Pods
    -0.24
    dealloc
    -0.24
     Outer
    -0.24
    ukt
    -0.24
    POSITIVE LOGITS
    ç³Ł
    0.29
    ivic
    0.26
     Section
    0.26
    èĢĮè¨Ģ
    0.26
    0.25
     деп
    0.25
     annex
    0.25
    梦è§ģ
    0.25
     section
    0.25
    0.25
    Activations Density 1.484%

    No Known Activations